Professional Documents
Culture Documents
Bronstein Move by Move - Steve Giddins
Bronstein Move by Move - Steve Giddins
by Steve Giddins
Contents
About this Publication
Acknowledgements & Bibliography
Series Foreword
David Bronstein: an Appreciation
1 Bronstein-Evans, Moscow 1955
2 Bronstein-Korchnoi, Moscow 1962
3 Bronstein-Winiwarter, Krems 1967
4 Nunn-Bronstein, Hastings 1975/76
5 Bronstein-Rojahn, Moscow 1956
5 Bronstein-Rojahn, Moscow 1956
6 Bronstein-Tal, Riga 1968
7 Bronstein-Ostrup, Gausdal 1994
8 C.Chandler-Bronstein, Gravesend 1996
9 Tal-Bronstein, Moscow 1966
10 Bronstein-Zamikhovsky, Leningrad 1970
11 Bronstein-Hunt, Maidstone 1994
12 Kaplan-Bronstein, Hastings 1975/76
13 Bronstein-Ljubojevic, Petropolis 1973
14 Bronstein-Ratner, Kiev 1939
15 Bronstein-Levenfish, Moscow 1949
16 Sax-Bronstein, Teesside 1975
17 Bronstein-Browne, Reykjavik 1990
18 Bronstein-Botvinnik, 22nd matchgame, Moscow 1951
19 Bronstein-Korchnoi, Moscow 1964
20 Bronstein-Keene, Teesside 1975
21 Bronstein-Najdorf, Budapest 1950
22 Bronstein-Szabo, Budapest 1950
23 Botvinnik-Bronstein, 17th matchgame, Moscow 1951
24 Bronstein-Keres, Gothenburg 1955
25 Bronstein-Golombek, Moscow 1956
26 Zita-Bronstein, Prague 1946
27 Pachman-Bronstein, Prague 1946
28 Reshevsky-Bronstein, Zurich 1953
29 Pein-Bronstein, London 1997
30 Langeweg-Bronstein, Beverwijk 1963
Back to Contents Page
Bronstein: Move by Move
David Bronstein is one of the greatest and most loved
chess players of all time. He won numerous major
tournaments and for many years he was one of the
world's strongest grandmasters. In 1951 he came
agonisingly close to winning the World Championship
title, drawing 12-12 against the reigning champion
Mikhail Botvinnik. Bronstein was one of the most
creative geniuses the chess world has ever seen, and
he left a legacy of wonderful games. In this book, FIDE
Master Steve Giddins selects and examines his
favourite Bronstein games, brings light to some games
which were previously unpublished, and shows us how
we can all learn and improve our chess by studying
Bronstein's play.
Move by Move provides an ideal platform to study
chess. By continually challenging the reader to answer
probing questions throughout the book, the Move by
Move format greatly encourages the learning and
practising of vital skills just as much as the traditional
assimilation of knowledge. Carefully selected questions
and answers are designed to keep you actively involved
and answers are designed to keep you actively involved
and allow you to monitor your progress as you learn.
This is an excellent way to improve your chess skills
and knowledge.
· Learn from the games of a chess legend
· Important ideas absorbed by continued practice
· Utilizes an ideal approach to chess study
Publisher Information
Acknowledgements
My thanks go, as ever, to John and Byron, for
commissioning the book, and then editing and
typesetting it. I am greatly indebted to Tony Stebbings,
Alan Hanreck and Rosemarie Hannan, who kindly
shared their memories of Bronstein with me, and also
supplied the original newspaper column in which the
score of Game 29 was published. Similarly, Cliff
Chandler supplied the score of his game with Bronstein,
which appears as Game 8. Ray Keene kindly lent me
his copy of the Toran/Eliskases book referred to in the
Bibliography, and he and John Nunn both confirmed
points relating to their own games against Bronstein
(Games 20 and 4 respectively). Finally, Matthew Sadler
educated me about the finer points of the Najdorf line
featured in Game 17.
The customary disclaimer of course applies.
Bibliography
D.Bronstein-K.Muchnik
5-minute game 1962
Ruy Lopez
(View in Game Format)
S.Giddins-D.Bronstein
British Rapidplay Championship, Leeds 1991
British Rapidplay Championship, Leeds 1991
Grünfeld Defence
(View in Game Format)
B.Olsson-D.Bronstein
Gausdal 1990
Gausdal 1990
Grob Opening
(View in Game Format)
Steve Giddins,
Steve Giddins,
Rochester, Kent,
February 2015
6
Trompowsky Idea
A45 XHGFEDCBAY
[Steve Giddins] 1RsNLmKQ+NtR!
1.d4 ♘f6 2.♗g5 ♘e4 3.♗f4 Diagram
2zPP+-+PzPP" 2
XHGFEDCBAY 3-+P+-+-+#
1RsNLmKQ+NtR! 4+-vLPzP-+-$
2zPPzPP+PzPP" 1 5-+-+-+-+%
3-+-+-+-+# 6+-sn-zp-+-&
4+-vLnzP-+-$ 7pzppzp-zppzp'
5-+-+-+-+% 8tr-vlkwqlsnr({
6+-+-+-+-& xhgfedcbay
7pzppzppzppzp' it looks as though White is a tempo up on the
Pirc line 1 e4 d6 2 d4 Nf6 3 f3, as he has
8tr-vlkwqlsnr({ gained the move Bf4, but Bronstein's clever
xhgfedcbay idea was to exploit this with 5...e5! 6.dxe5
Diagram
and now Bronstein's suggestion was d6
. After 4.f3 ♘f6 5.e4 Diagram
XHGFEDCBAY
[ Alan later got the chance to play this 1RsNLmKQ+NtR!
novelty against 'Tromp' specialist, GM 2zPP+-+PzPP" 3
Julian Hodgson, at the 4NCL in 1996.
Hodgson preferred 5.♘c3 , but after ♘bd7 3-+P+-+-+#
6.e4 e5 7.♗e3 ♗e7 8.♘ge2 c6
Black had a reasonable Philidor position,
4+-vLP+-+-$
and went on to beat his powerful GM 5-+-zP-+-+%
opponent. ]
6+-sn-zp-+-&
7pzpp+-zppzp'
(Diagram 2)
8tr-vlkwqlsnr({
xhgfedcbay
♘h5! . Now 7.♗e3
[ and 7.♕d2 ♘xf4 8.♕xf4 ♘d7!
regains the pawn, or else offers excellent
dark-square compensation after 9.exd6?!
♗xd6 ]
(Diagram 5)
XHGFEDCBAY 9
Olsson,B
A00
1-mK-+-wQ-+![ Bronstein,D
2+PzP-+-+P" 6 Gausdal 1990
[Steve Giddins]
3P+l+-+P+#
1.g4 d5 2.♗g2 ♗xg4 3.c4 c6 4.cxd5 cxd5
4+-+-tR-zp-$ 5.♕b3 ♘f6 6.♕xb7 ♘bd7 7.d4 ♖b8!
5-+-vL-+-zp% 8.♕xa7 ♕c8 9.♗f4 e5 10.dxe5 ♗c5
11.♕a4 ♖b4 12.♕c2 ♖xf4 13.exf6 ♗xf2+
6sN-+-zp-+-& 14.♔d1 ♖c4 15.fxg7 ♖g8 16.♘c3 ♖d4+
7pzpp+q+-+' 17.♔c1 ♗e3+ 18.♔b1 ♖d2 19.♕xh7 ♘f6
20.♕h4 d4 21.♕xf6 ♖xb2+ 22.♔xb2
8mk-vl-tr-+-( ♕xc3+
xhgfedcbay 0-1
16.♖c1 and the c4-pawn is doomed. ] , with a winning attack. Yet again, the
13.♗b2 ♗g7 black king is caught in the centre, with
[ QUESTION: Can't Black take the pawn devastating effect. ]
with 13...♘xb4 , attacking the bishop on [ EXERCISE: How can White prevent the
c2? escape of the black king after 15...♕e7 ? ]
ANSWER: Yes, but after the simple reply
14.♗b1 , he still has serious problems. The
pawn on e5 is hanging and cannot be
XHGFEDCBAY
defended in any comfort: for example, ♘c6 1-mKR+QtR-+![
15.a4 b4 16.♕c2 , pinning the knight, and 2zPPzP-sN-vLP" 11
threatening both 17 d5 and 17 dxe5. Black
has no satisfactory reply, and will 3-+N+-+L+#
inevitably lose back the e5-pawn, with
his kingside still undeveloped and his
4+-+PzP-zP-$
king exposed in the centre. ] 5-+-zp-+p+%
Evans tries to avoid this by hastening to
remove his king from the centre, but his 6+psn-zpn+p&
problems remain. 7pvlpwq-+l+'
14.♖c1 ♗b7 15.♗b3 Diagram
XHGFEDCBAY 8tr-+k+-+r(
1-mKR+QtR-+! xhgfedcbay
2zPPzP-sN-vLP" 10 It looks as though Black has managed to
shore things up, and is now ready to castle.
3-+N+-+L+# ANSWER: 16.♖xc6! This exchange sacrifice
cries out to be played, but, as we will see,
4+-+PzP-zP-$ Bronstein had to calculate quite a long way,
5-+-zp-+p+% in order to prove that it works.
♗xc6 17.dxe5 ♘h5 Diagram
6+psn-zpn+p& [ The first point is that 17...dxe5? 18.♘xe5
7pvlp+-wql+' attacks both c6 and f7, and is obviously a
disaster for Black. His 17th move is
8tr-+k+-+r({ th eref ore his only hope, and, indeed,
xhgfedcbay Bronstein even goes so far as to suggest
that Evans may have seen this far and relied
♕e7 Diagram on 17...Nh5, when playing his 10th move. I
[ 15...♕b8 is the alternative, but then White am not so sure about that, but, in any event,
h a s t h e c o m b i n a t i o n 16.dxe5 ♘xe5 Bronstein certainly had to consider the
( 16...dxe5 17.♖xc6 ♗xc6 18.♘xe5 consequences of this move, before
is even worse than the game) 17.♗xe5 sacrificing the exchange. ]
dxe5 18.♗xf7+ ♔xf7 19.♘g5+ ♔e7
20.♕b3 ♕g8 21.♖c7+ ♘d7 22.♕d3
[ Black has no good way to protect the f6- trouble at this point, and so the rest
pawn, since 26...♔g7 is met by 27.g5 . ] happened at great speed. In any event,
27.g5! Even better than taking on e5. White's position is overwhelming and he
QUESTION: Reti would have approved! duly wraps up the full point. ]
ANSWER: Yes, the Hypermoderns would all 34...♖xe6 35.♖xe6 ♗e8 36.♖e7 ♕c1
have been delighted with the way the white 37.♖xe8+ ♔xe8 38.♕e4+ ♔d8 39.♕e7+
queen operates from the corner of the board, ♔c8 40.♗e6+
and t h e g- pawn un der mi n es t h e en emy [ 40.d7+ mates a couple of moves quicker,
fortifications on the long black diagonal. but who is really counting at this point? ]
♖e8 40...♔b8 41.♕c7+
[ Bronstein's criticism of his earlier play [ Here too, 41.d7+ mates in a few moves. ]
was based on the claim that here, Black 41...♕xc7 42.dxc7+ ♔b7 43.♗xf7 ♖f8
could have made White's task far from 44.♗d5+
s i m p l e w i t h 27...♕xd6 28.gxf6 ♖e8 1-0
. However, the computer shows that this
does not really help, and White crashes
t h r o u g h a f t e r 29.f7! ♖e7 30.♖c1! 11 C83
. The threats of Rxc6 and Rc5 are Bronstein,D
unstoppable, and Black is tied hand and Korchnoi,V
foot. ( Instead, Bronstein only considered Leningrad-Moscow m., Leningrad (Game 2)
t h e i n f e r i o r 30.♗xe5 ♕xe5 31.♕xe5 [Steve Giddins]
♖xe5 32.♖c1 ♗b7 33.f3 ♖e7
, which is indeed still not fully clear. )] 1.e4 e5 2.♘f3 ♘c6 3.♗b5 a6 4.♗a4 ♘f6
28.♕d4 5.0-0 ♘xe4 The Open Spanish, which has
[ 28.gxf6 ♘f7 29.♕d4 was equally good. ] always been a favourite of Korchnoi. He was
28...♕g4 29.f3 ♕d7 30.gxf6 ♘f7 31.e5 introduced to it by the legendary Leningrad
The position is picturesque in the extreme. trainer, Vladimir Zak, who was always a
The white pawns are advancing down the great lover of the line.
board, like Space Invaders, crushing all 6.d4 b5
befo re th em. The g ame p r es en ts a nice [ Q U E S T I O N : T h e m o v e 6...exd4
parallel with Game 5 against Rojahn. loses by force, doesn't it?
g5 ANSWER: Actually, no. There are
[ 31...♘d8 attempts to hold up the e-pawn, considerable complications after 7.♖e1
but then, for example, 32.♘e4 followed by , but it is not clear that Black is so badly
Nc5 or Ng5 is crushing. ] off, strangely. The main line is d5 8.♘xd4
32.♖e1 gxh4 33.e6 ♕c8 34.♕xh4 ♗d6 9.♘xc6 ♗xh2+ 10.♔h1! ♕h4
[ The human move, but the computer points 11.♖xe4+! dxe4 12.♕d8+ ♕xd8
out that 34.d7 is actually mate in 8, 13.♘xd8+ ♔xd8 14.♔xh2 , with an
improbable as this may seem. ] endgame where White has two pieces for
[ Bronstein also points out that 34.♕g4 rook and two pawns. It has always been
"would have forced immediate resignation", regarded as better for White, but this is
but both players were in serious time- not 100% clear. Unfortunately, though,
Black's main problem is that he is only crushes Black on the dark squares,
likely to draw at best, which makes it less exploiting the hole on c5. Such a scenario
attractive. ] is only possible if Black mishandles it, but
7.♗b3 d5 8.dxe5 ♗e6 9.c3 it is a frequent threat. ]
[ 9.♕e2 , ] 9...♗e7
[ 9.♘bd2 , ] [ This old line is somewhat discredited
[ and 9.♗e3 are all possible, ] these days, with the more active moves
[ but 9.c3 has always been the main line. 9...♘c5 ]
QUESTION: And what essentially is going [ and 9...♗c5 both enjoying a better
on here? Black's position looks quite reputation. ]
attractive – he has free play, a strong 10.♗c2 0-0 11.♕e2
knight on e4 and a queenside pawn [ 11.♘bd2! is more precise. ]
majority. Why shouldn't he have a great 11...f5
position? [ In his notes in 'David against Goliath',
ANSWER: Dr. Tarrasch would be proud of Bronstein points out that the players had
you! The great German player always reached the same position just two days
argued just as you do, claiming this was earlier, at the Soviet Team Championships.
the only correct defence to the Spanish, T h e r e K o r c h n o i h a d c h o s e n 11...♘c5
but most great masters have preferred and the game had been drawn. ]
White here. 12.exf6 ♗xf6? This pawn sacrifice was
QUESTION: Why? obviously Korchnoi's idea, but it proves
ANSWER: The problem is that, when one insufficient.
looks a little more deeply, things are not [ Instead, Black could equalize with
quite so wonderful for Black. Firstly, his 12...♖xf6 : for example, 13.♘g5 ♗c5!
knight looks great on e4, but is not "with very interesting complications"
entirely stable – in some lines, White can (Bronstein). The computer supports this
play c4, exploiting the overloading of the view, one variation running 14.♘xe4 dxe4
d5-pawn, which must defend e4, and in 15.♕xe4 ( 15.♗xe4? ♗c4 ) 15...♗f5
other variations, White can kick the 16.♗b3+ ♔h8 17.♕h4 ♗xb1 18.♖xb1
knight with f3. Secondly, the white pawn ♖xf2! 19.♕xd8+ ♖xd8 20.♖xf2 ♖f8
on e5 means that Black must always be and Black regains his pawn, with equal
careful to ensure that his king does not chances. ]
come under attack. And playing ...f5 or ... 13.♘bd2! This is the point. White exploits
f6 himself weakens the kingside further, the pin on the undefended bishop at e6, to
as well as the diagonal a2-g8. Finally, the win a pawn.
black queenside pawns have advanced [ Instead, 13.♗xe4? dxe4 14.♕xe4 ♗d5
rather prematurely, and there is a danger would be much better for Black. ]
that the c-pawn will end up backward. 13...♗f5
Capablanca, in 'Chess Fundamentals', gives [ 13...♘c5 is met by 14.b4 ♘xb4 15.cxb4
an example of a nightmare scenario for ♗xa1 16.bxc5 with a clear advantage. The
Black, where White plays Nd4, forces the two knights would be much stronger than
exchange, recaptures cxd4, and then the rook in this position.
but with plenty of scope to outplay the temporary state of affairs. Meanwhile, White
opponent, ( as 9.exd5?! e4 10.dxc6 ♗g4 sets up a solid central structure, and will now
is quite dangerous. )] turn to the exploitation of the offside bishop
8.c3 ♗g4?! As discussed in Game 1, this on h5. This will be driven back to g6, by an
move is somewhat dubious before White has eventual g2-g4, and only then will White
advanced his pawn to d4. expand in the centre. White's advantage is
[ The usual move is 8...0-0 with equal not a large one, but it is stable and long-
chances. ] lasting. There have been many games in
9.h3 Diagram such structures, and these have shown that
XHGFEDCBAY Black suffers from his queen's bishop being
stuck out of play on g6.
1-mKR+-vLNtR! 0-0 11.♘bd2 ♘a5 12.♗c2 c5 13.♖e1
2+PzPQzP-zPP" 14 All thematic stuff from both sides. With his
last move, White freed the f1-square for his
3P+N+-zPL+# knight, and now intends Nf1, g4 and Ng3.
♘d7 Diagram
4+l+P+-+-$
5-+-zp-+p+% XHGFEDCBAY
6+-sn-zpn+p& 1-mK-tR-vL-tR![
7pzppvl-zp-+' 2+PzPQsNLzPP" 15
8tr-+kwq-+r({ 3P+N+PzP-+#
xhgfedcbay 4+-+P+-+-$
5l+-zp-zppsn%
♗h5!? QUESTION: The punctuation
suggests you are not entirely convinced by 6+-+-zp-+p&
this. 7pzppvln+-+'
ANSWER: True. The move is consistent, but
now the bishop is liable to end up out of play 8+ktr-wq-+r(
for some time to come. On the other hand,
retreating back along the c8-h3 diagonal, for
xhgfedcbay
e x a m p l e t o d 7 , w o u l d o b v i o u s l y b e a 14.g4! QUESTION: What is the idea of that?
significant concession, as White would have [ ANSWER: I think Winiwarter intended to
gained the useful move h2-h3 for nothing. meet 14.♘f1 with ♗xf3 15.♕xf3 ♗g5
10.d3! QUESTION: An exclamation mark , seeking to relieve his somewhat
seems a bit much for such a modest move. cramped position by a couple of exchanges
Hasn't Black achieved something, by forcing and, in particular, to offload his rather
White to abandon his planned central bad dark-squared bishop, which is hemmed
advance d2-d4? in by the black central pawns. However,
ANSWER: He has delayed it for the time Bronstein does not allow this. ]
being, but as we will see, this is only a 14...♗g6 15.♘f1 f6?! QUESTION: That
doesn't do much for his king's bishop! and White would be much better after
ANSWER: No. If Winiwarter had followed 19.♘xd4 ) 19.a4 White retains the typical
Jonathan Rowson's advice and "talked to his Lopez pull. ]
pieces", I suspect the king's bishop would 18.d5 QUESTION: That's a major decision,
have had some choice words to say about the isn't it?
move 15...f6. ANSWER: Yes, it is. White abandons any
QUESTION: So what do you recommend? attempt to break through in the centre or
[ ANSWER: I think 15...h6! is better, still exploit the weakened d5-square, and instead
hoping to trade off his bad bishop with a places his hopes on a flank breakthrough. In
subsequent ...Bg5. White would retain fact, as will become clear later on , his
some advantage, but less than in the strategy is very deep. Nowadays, thanks
game. particularly to Karpov's influence, d4-d5 is
QUESTION: But is White better, even in pretty much the standard approach to these
the game? Fritz does not seem to think so. Spanish structures. Fischer was always
ANSWER: Yes, I noted that also. But I rather more keen on the Rauzer plan with
think this is one case where the dxc5 and trying to exploit the d5-square, but
computer's judgement must be taken with a Karpov perfected the slow manoeuvring
pinch of salt. Computers still have a lot of approach to the structure.
trouble assessing such blocked and slow c4 19.b4 ♘b7 See the note to move 15.
positions. I repeat that White's advantage Now the knight is very bad on b7, which is a
is not enormous, but he is definitely for typical problem for Black in these positions.
preference here. Black's queen's bishop is QUESTION: So why doesn't Black take on
out of play, his king's bishop is bad, and b3?
once the white pawn reaches d5, as it will [ ANSWER: The cure would be no better
inevitably do, Black will have the usual than the disease. After 19...cxb3 20.axb3
Chigorin system problems with his queen's White has pressure down the a-file against
knight. ] the pawn on a6, whilst the knight still
16.♘e3 ♗f7 17.d4 Here we go. As predicted, lacks any decent retreat square from its
White's restrained set-up with the pawn on current offside position on a5. White can
d3 was only a temporary expedient, and now continue with Bd2, setting up tactical
that he has completed the necessary threats against the knight, with c3-c4. ]
preparations, he expands in the centre. 20.a4 ♕c7 Diagram
♖e8 QUESTION: Could he exchange in the
centre? (Diagram 16)
[ ANSWER: It is also possible, yes, but
after 17...cxd4 18.cxd4 ♖c8 ( capturing
twice in these structures, with 18...exd4
is rarely good, unless Black can follow up
with a quick ...d6-d5 and/or exert pressure
against the white centre, with moves such
as ...Bb7, ...Re8, etc; here, his pieces are
all hopelessly misplaced for such a strategy,
pawns gone, and White's pawns moving up to ♗f7 33.♗d1 ♗e8 34.♗e2 ♖c8 35.♘e3
b4, c4 and then c5, one can see that Black's The last brick in the wall. Note how White
position will be extremely hard to defend. left this move until last, maintaining the
Th e acute lack o f s pace f or h is p ieces , knight on f5 as long as possible, where it
especially the way the knight on b7 is under attacked the h6-pawn and so prevented
its colleagues' feet, makes it very hard to Black bringing his rook on h7 over to help
defend. on the queenside.
g5 26.♘f5+ ♔f8 27.♗e3 ♖h7 28.♘d2 ♘b8 Diagram
Diagram XHGFEDCBAY
XHGFEDCBAY 1-+Q+-tR-tR![
1R+-+-+-tR! 2+KvLLsN-+-" 18
2+KzPQsNL+-" 17 3-+PsN-zP-+#
3-+-vL-zP-+# 4+P+P+pzP-$
4+P+P+pzP-$ 5Pzp-zpP+pzP%
5PzpNzpP+pzP% 6zp-zp-zp-+p&
6zp-zp-zp-+p& 7r+-+-wqn+'
7r+lvlnwqn+' 8+-mklvlrsn-(
8+-mk-+-+r({ xhgfedcbay
xhgfedcbay 36.♘dxc4! The breakthrough finally comes.
White's next stage is to bring his pieces into Black may as well capture, else the knight
position for the sacrifice on c4. He proceeds just moves away again, and White plays c4
to do this systematically. without even having invested any material.
♗g8 29.f3 ♗d8 30.♗f2 ♗e7 QUESTION: bxc4 37.♘xc4 ♗b5 Losing the exchange,
Black just seems to be shuffling his pieces but even the computer confirms that the
aimlessly. position is hopeless for Black.
ANSWER: Yes, because he has no [ A f t e r a m o v e s u c h a s 37...♕f7
constructive moves at all. He can only await , for example, meeting the threat of Nb6,
the blow falling and then hope to be able to W h i t e c a n p l a y s i m p l y 38.♘e3
deal with the tactics that result. . Now a6 hangs, and cannot be defended.
31.♖hc1 ♗d8 32.♕f1 In contrast with his Once that goes, the white queenside pawns
opponent's, White's play is extremely will be unstoppable. Note how the black
systematic. His rook stands behind the c- pieces are helplessly stumbling over each
pawn, ready to support an eventual c4-c5. other – the knight on b7 does not have a
Now he just needs to transfer his Spanish single move, for example. ]
bishop to e2 and bring back the knight from 38.♘b6 ♗xe2 39.♕xe2 ♗e7 40.♘xc8
f5 to e3, and all will be set. ♕xc8 41.♗a7 The simplest. White does not
14 C58
Bronstein,D
Rojahn,E
Moscow Olympiad (Game 5) 1956
[Steve Giddins]
the Giuoco Piano, a well-known tabiya is ♖xa2 35.♖e7 ♖a4 36.♖e2 ♖xd4
reached after 3...♗c5 4.c3 ♘f6 5.d4 exd4 37.♔f2 a4 38.♖a2 ♔e4 39.♔e2 f4
6.cxd4 ♗b4+ 7.♗d2 ♗xd2+ 8.♘bxd2 d5 0-1, E.Pogosyants-D.Bronstein) 12...♖xf7
9.exd5 ♘xd5 10.♕b3 . is then critical, when Bronstein claims
A) Now the best move is generally agreed that Black has sufficient initiative to
to be 10...♘a5 , but this has the compensate for the pawn. That is
significant drawback that White can force perhaps objectively questionable,
a d r a w a f t e r 11.♕a4+ ♘c6 12.♕b3 although Black's position does not looks
, etc.; so much worse than in the main line of
B) Black players wishing to fight for a the 4 Ng5 Two Knights.;
win here are therefore forced to resort to D) 10...-- ]
10...♘ce7 , which avoids the immediate 4.♘g5 QUESTION: Is this good? Tarrasch
draw, although White probably retains a dismissed it as bad, didn't he?
pull, objectively.; ANSWER: Yes, Dr. Tarrasch famously
C) I for one certainly never seriously derided it as "a duffer's move", whereas
entertained the thought that Black might Steinitz stubbornly stuck to it, despite losing
have any other playable move in this a number of bad games against Chigorin.
position, but I had failed to reckon on the The dispute rumbled on for decades, but the
tireless creativity of David Bronstein. In advent of computers has certainly swung
a clock simultaneous in the 1960s, against grandmaster opinion towards the white side,
the well-known endgame study composer and nowadays, one finds few GMs willing to
Ernest Pogosyants, Bronstein played the play the black side of the Two Knights.
remarkable move 10...0-0!? . d5 5.exd5 ♘a5 And this is the main reason
It looks like madness, but, on closer why. Not only does Black lose a pawn, but he
i n s p e c t i o n , a f t e r 11.♗xd5 ♘a5 is also forced to put his knight on the edge of
B l a c k r e g a i n s t h e p i e c e . 12.♗xf7+ the board. Although White is behind in
( in the game, Pogosyants refrained from d e v e l o p me n t , h i s p o s i t i o n c o n t a i n s n o
taking the pawn, and was holding the weaknesses at all, and nowadays, strong
balance until very near the end; the rest players tend not to trust such sacrifices.
of the game is not of great importance in [ Unfortunately for Black, the two
itself, but once again, I offer the alternative continuations 5...b5 ]
complete score, for the benefit of [ and 5...♘d4 , which see the black knight
collectors of Bronstein game scores: occupy a central post, have both been
12.♕c3 ♕xd5 13.0-0 ♘c6 14.♖fe1 analysed out to a white advantage. ]
♗e6 15.♘b3 ♕h5 16.♘e5 ♗d5 6.d3?! A very unusual choice for the 20th
17.♘xc6 bxc6 18.♕g3 a5 19.♘c5 ♖fe8 century.
20.♖e5 ♕g6 21.♕xg6 hxg6 22.♖ae1 [ The line that has put people off the Two
♖xe5 23.♖xe5 ♖b8 24.h4 ♖xb2 K n i g h t s i s 6.♗b5+ c6 7.dxc6 bxc6
25.♖e8+ ♔h7 26.♘d7 g5 27.hxg5 a n d n o w b o t h 8.♕f3 ( and Steinitz's old
♔g6 28.♘e5+ ♔xg5 29.f4+ ♔f6 move 8.♗d3 have achieved excellent
30.♘g4+ ♔f5 31.♘e3+ ♔xf4 results for White in the computer era. )]
32.♘xd5+ cxd5 33.♖e7 f5 34.♖xg7 [ QUESTION: But is there anything wrong
w i t h B r o n s t e i n ' s 6.d3 ? It looks quite QUESTION: But his opponent obviously did
logical to defend the bishop and open the not know about this?
other white bishop's line. ANSWER: No, According to Bronstein, after
ANSWER: It is, but the problem is the the game, his opponent complimented him on
sequence which follows. ] his coolness and cold-bloodedness, in calmly
6...h6 7.♘f3 e4 This is the rub. playing on and finding the best chance, after
[ For many decades, it has been assumed "blundering" away a piece!
t h a t a f t e r 7...e4 White has nothing better ♘xc4 9.♕d4 ♘b6?! This is already a
than 8.♕e2 , but after ♘xc4 9.dxc4 ♗c5 critical moment.
Black has excellent play for the pawn. [ Rojahn's choice is the most obvious, but
However, for this game, Bronstein had in actual fact, 9...♘d6 was more critical.
prepared a very special idea. ] QUESTION: But that just sets the knights
8.dxe4!? Astounding. up to be forked, doesn't it?
QUESTION: Err, doesn't this...? ANSWER: It looks that way, but tactical
ANSWER: Yes, it does. White is giving up a resources change the assessment, and after
whole piece, for which he will have only a 10.♘c3 Black has the nasty trick ♘fxe4
couple of pawns, plus a central pawn duo. At 11.♘xe4 ♕e7 , when he regains the piece
first sight, it looks like madness, and indeed, and eliminates one of White's central pawns.
the line has never been trusted and has not After the further moves 12.0-0 ( 12.♘e5
been repeated at GM level, but it works on ♘xe4 13.♕xe4 f6 14.f4 fxe5 15.fxe5
this occasion. ♕b4+ is also better for Black) 12...♘xe4
QUESTION: What was the reaction of ( and not 12...♕xe4?? 13.♖e1 ) 13.♖e1 f5
Bronstein's teammates? it looks a bit shaky, but Black should
ANSWER: Ah, well thereby hangs an emerge with most of his booty intact: for
interesting tale. This game was played in an example, 14.♘d2 ♕c5 15.♕e5+ ♔f7
Olympiad. The Soviets had a very good rule 16.♘xe4 fxe4 17.♕xe4 ♗d7
in operation within their team, to the effect and, objectively, White does not have
that nobody was allowed to play in a manner enough.
that mig ht caus e their teammates to b e QUESTION: So is 9...Nd6 just a
alarmed about their position. This was for refutation? What did Bronstein intend?
the very good pragmatic reason that, in a ANSWER: I am not sure! It is curious
four-board match, if the other players start to that neither in '200 Open Games', nor
believe that teammate X is losing, they are 'The Sorcerer's Apprentice', does
liable to start taking liberties in their own Bronstein mention the move 9...Nd6, or say
games, to try to generate winning chances to what he had in mind. ]
make up for him. This very often makes the 10.c4 c5? Black was clearly totally thrown
match situation worse, and so the Soviets by events on the board, and it is perhaps not
al wa y s t r i ed t o av o i d s u ch d an g e r s . I n surprising that he found it hard to orient
accordance with this 'rule', Bronstein records himself. The mass of white pawns from c4 to
that he had shown his idea to his teammates e4 looks dangerous, and Rojahn hastens to
the night before, and had received their blockade the c-pawn, but the result is that
blessing to his trying it out. the white d-pawn is now passed, whilst
control of e5. The computer still assesses the [ 24...♘xe5 was a try to disrupt the course
position in Black's favour. of White's build-up. After 25.fxe5 ♕xf3
19.♖ae1 ♗xc3 20.♕xc3 ♕f6 ( or 25...♕xe5 26.♘xc5 when "White has
[ 20...f6 is the computer's choice here, but an extra passed pawn" – Vainstein) 26.♖xf3
after 21.f4 play continues along similar ♗xe4 27.♖xe4 "the double rook ending
lines. Fritz still claims a small edge for with an extra pawn is won for White"
Black, but its assessment is dropping (Vainstein). ]
with every move, which usually means that [ 24...♗h5 is a better option, also mentioned
the players knew what they were doing. ] b y V a i n s t e i n , w h o t h e n g i v e s 25.♘g3
21.e5 ( Fritz's simple line 25.♕e3 ♗g6 26.♘d6
[ QUESTION: Wait a minute! Isn't 21.♕h3 ♕d3 27.f5 ♕xe3+ 28.♖xe3 looks even
attacking two pieces? more crushing) 25...♗xf3 26.♘xf5 ♗g4
ANSWER: Yes, but Rojahn had seen this 27.♘e7+ ♔h8 28.f5 and claims a winning
and intended the reply ♕b2 . If then advantage for White. ]
22.♘b1 ♗e2 23.♕xd7 ♗xf1 24.♖xf1 [ Finally, there is 24...♕h5 , which the
♕xa2 and Black has returned the exchange, computer suggests as the best chance. After
to neutralize the white initiative. 25.♕xh5 ( 25.g4 ♗xe4 26.♖xe4 ♕g6
QUESTION: So you mean Black was 27.f5 ♕g5 is less clear, according to the
trying to equalize? beast ) 25...♗xh5 26.f5 White retains what
ANSWER: So it would seem, and should be a decisive advantage, although
Bronstein's response shows that he was some work remains to be done. ]
not interested in such a continuation. 25.g4 ♕g6 26.f5 Now it turns into a
Although Fritz still thinks Black is better, massacre, as the pawn armada sweeps away
it is clear that both players were of a all resistance.
different opinion during the game. ] ♕b6 27.♕g3 f6 28.e6 If anything justifies
21...♕f5 22.f4 Now the pawn centre really Bronstein's original idea of 8 dxe4, it is this
starts to assume menacing proportions, and position.
even the silicon beast agrees that White is ♘e5 29.h4! It seems only fair to include
now very much on top. Bronstein's faith in White's eighth and final pawn in the fun.
his position has been justified, and the ♔h8 30.g5 ♖bc8 31.♔h1 QUESTION:
remainder of the game is pure pleasure for Huh?
White. ANSWER: Yes, I know it looks a bit strange.
♗g6 23.♘e4 ♖ab8 QUESTION: What is But besides setting up gxf6 and then Rg1,
the point of that? there is probably also an element of sadism
ANSWER: Black just defends the b7-pawn about it – Black is invited to consider that,
against the threat of 24 Nd6. But such a if White has time for such moves, perhaps it
passive move can hardly be good. is worth considering resignation.
[ 23...b6 looks like a better way to do the ♕d8 Not on your life!
same thing, although it does not change 32.g6 ♗xg6
the course of the game. ] [ Unable to face the horrors of 32...♗g8
24.♕f3 Now g4 is also on the agenda. 33.d6 Black makes a desperate sacrifice,
♗h7 but it does not long delay the inevitable. ]
ANSWER: Then White can continue 8.♗e3 move ago, since now d5 is defended as
disposing of the threat against f2. Once well. ]
again, we quote John Shaw's book: ♕xd5 9.♗e3 ♘xc3 QUESTION: Is this forced?
9.♗xc5 ♕xc5 10.♘c3 0-0 11.♘xe4 fxe4 [ ANSWER: No. Black has a choice of two
and now White again implements his similar endgames here. The other option is
strategy of heading for the better endgame: 9...♗xe3 10.♕xe3 ♘xc3 11.♕xe7+ ♔xe7
12.♕c4+! ♕xc4 13.♗xc4+ ♔h8 14.♘e5 12.bxc3 .
and he is just better, with the more active QUESTION: And how is this position?
pieces, better development and fewer ANSWER: Once again, White is better. For
pawn islands. ] the moment, he is a pawn up. Black can
regain the pawn on c2, but after ♗xc2
XHGFEDCBAY ( the main line is 12...♗e4 13.♘g5 ♗xd5
, but here, too after 14.0-0-0 , White's
1R+LmK-vLNtR![ superior development and initiative gives
2zPP+Q+PzPP" 27 Black problems; again, you should consult
a book on the King's Gambit for more
3-+N+-+-+# details, but for now, I ask you to trust me)
13.♔d2 ♗g6 14.♖e1+ ♔d6 15.♘d4
4+-zPn+-+-$ White has a strong initiative. ]
5-+l+Pvl-+% Tal chooses another sequence, also leading to
an ending where White will have the
6+-+-+-+-& superior development and some initiative.
7pzpp+-zppzp' 10.♗xc5 ♘xe2
[ 10...♕xe2+ 11.♗xe2 ♘xd5 also regains
8tr-+kwq-snr( the pawn, but then the further pawn
xhgfedcbay sacrifice 12.0-0-0! ♘xf4 13.♖he1
once again repeats the familiar scenario –
8.♘c3! White is fully mobilized, whilst Black has
[ QUESTION: Wait a minute. Isn't 8.g4 most of his pieces at home, and his king
winning a piece here? is exposed in the centre.
ANSWER: Yes, but it has been known QUESTION: It seems strange that king
since the 1920s that Black is much better safety should be such an issue, even with
after 0-0 9.gxf5 ♖e8 , as was the queens off the board.
demonstrated by a famous game ANSWER: Well, although I have been
Spielmann-Tarrasch, Mahrisch-Ostrau referring to these various positions as
1923. If you do not believe me, John "endgames", they are more in the nature
Shaw's book is again the place to look for of queenless middlegames. With so many
detailed analysis, as space considerations open lines and a large development
preclude us from going into the variations advantage for White, the traditional
more fully here. ] middlegame factor of king safety is still
8...♕e7 important, even in the absence of the
[ 8...♗f2+? 9.♔d1 is even worse than a ladies. ]
11.♗xe7 ♘xf4 12.♗a3 QUESTION: We but the computer suggests the simple 14.d6
have another, very similar position to the last i s v i r t u a l l y w i n n i n g f o r W h i t e : ♗xf3
note! 15.gxf3 c6 ( or 15...cxd6 16.♗xd6 ♘e6
ANSWER: Yes. As we have seen, there are a 17.f4 and the two bishops slice Black's
number of similar queenless positions which position apart) 16.♖e1+ ♘e6 17.♗h3
can arise from this line, and in all cases, the , with a large advantage. ]
key factors are the same. Black can regain 14...♗xd5 15.g3 This is the point of
his gambit pawn, and even nick a further Bronstein's idea. He was evidently
pawn, if he wishes, but White has a convinced that it was best, since he offers no
significant lead in development, plus two comment on it in '200 Open Games', but the
strong bishops, which are generally enough computer is sceptical.
to promise him the better chances. In fact, as ♗xh1 16.gxf4 QUESTION: What has White
Bronstein acknowledges in '200 Open Games', obtained for the exchange?
this very position was all old analysis by ANSWER: He has a fierce initiative, with
Tartakower, in his 1920s book, Bc4 coming next. But it looks as though
'Hypermodern Chess'. Black could have returned the exchange in
♘d7 order to equalize.
[ Tal decides against larceny, either by c5?!
12...♘xd5 13.0-0-0 , ] [ 16...0-0-0 is Fritz's choice, offering back
[ or 12...♗xc2 13.♖c1 . ] the exchange, to get the king out of trouble.
13.0-0-0 ♗e4 After 17.♘xf7 ♖he8 18.♘xd8 ♔xd8
[ QUESTION: What is wrong with it appears that White does not have very
13...0-0-0 here, getting the king out of much. ]
Dodge while he has the chance? 17.♗c4 ♗c6 18.♘xf7 Now the game reverts
ANSWER: Amazingly, it loses a whole to the lines Bronstein had envisaged. Despite
piece! the simplified position, White has a very
EXERCISE: How? powerful attack, which more than
ANSWER: 14.♖d4 and wherever the compensates for the exchange.
knight moves, Black loses a piece: ♘g6 b5 QUESTION: Is there anything else?
( or 14...♘h5 15.g4 ) 15.g4 and the bishop [ ANSWER: Not really. The only other
on f5 is trapped in broad daylight! An old option worth considering is to chase the
friend of mine, a strong county player, bishop by means of 18...♘b6 , but then
once fell into a similar trap against the 19.♖e1+ ♔d7 20.♗e6+ ♔c7 21.♗xc5
veteran English master Stuart Milner-Barry, and the threat of 22 Bd6+ forces Black to
in a tournament in the late 1960s. As if surrender the exchange anyway. After
losing a piece in such a fashion were not ♖hd8 22.♘xd8 ♖xd8 23.f5 White has an
embarrassing enough, he then had to see extra pawn, the two bishops and still a
the game published in the BCM, with notes strong initiative, and should win without
by Harry Golombek, who described my any great trouble. ]
friend's play as "mildewed". ] 19.♘d6+?! Another inaccuracy, which
14.♘g5?! imperils White's advantage.
[ The start of an imaginative sacrificial line, [ The computer points out that simply
19.♗b3! retains all White's threats, since c4 does not have the time to take on h2 or f4.
20.♗xc4 bxc4 21.♘xh8 is much better for The knight cannot move because of mate on
him. ] h7, so Black is driven to desperate measures.
19...♔e7 20.♘xb5 ♖hf8? Now White is c4 33.♗f5 c3
again winning. [ Black is losing a piece, in view of 33...♖e8
[ Black should have preferred 20...♗xb5 34.♗xf6 . ]
. Bronstein considered this, and gave the 34.♗xc8 cxb2+ 35.♔xb2 ♖xh2 36.♖xa7
variation 21.♗xb5 ♖hd8 22.f5 ♖ac8 White has regained his exchange and has no
23.b3 , concluding that "the pressure fewer than three extra pawns.
ex e r t ed b y Wh it e's b is h o p s wo u ld b e ♖f2 37.♖a4 ♔g6 38.♖d4 h5 39.a4 h4
extremely unpleasant". This is perhaps 40.a5 ♗g2 41.a6 ♘h5 42.♗b7 ♘xf4
typical of pre-computer assessments. On 43.♖xf4
the face of it, his conclusion seems [ After 43.♖xf4 ♖xf4 44.♗xg2 ♔xg5
perfectly valid, but Fritz just dismisses the 45.a7 the a-pawn costs Black a rook. It is a
position as equal after, for example, ♘f6 . ] pity that two errors could have cost
21.♘d4! ♗g2 22.♘e6 ♖f5 23.♖g1 ♗e4 Bronstein his advantage, but overall, this
24.♘c7 was an important game, both from a
[ Good enough, but Bronstein himself theoretical viewpo int, and als o as an
pointed out that 24.♖e1! wins at once. He illustration of the power of the initiative
admitted that he simply overlooked that, in queenless middlegames. ]
after ♘f6 White has 25.♘xg7 , winning: 1-0
for example, ♖xf4 26.♘e6 ♖h4 27.♘xc5
♔d8 28.♘xe4 ♘xe4 29.♗d5 , etc. ]
24...♖d8 25.♖xg7+ ♔f6 26.♖f7+ 16 C39
Bronstein admitted that he actually thought Bronstein,D
this was mate, and had missed the fact that Ostrup
the black king can flee to g6! Gausdal (Game 7) 1994
[ H e i s s t i l l w i n n i n g , b u t 26.♖xh7 [Steve Giddins]
was simpler. ]
26...♔g6 27.♖e7 QUESTION: Is it clear that 1.e4 e5 2.f4 exf4 3.♘f3 g5
White is winning? He has two pawns for the [ Bronstein's lifelong love affair with the
exchange, but f4 and h2 both look weak. King's Gambit extended back to the
ANSWER: The decisive factor is still the earliest days of his chess career. In the
strength of White's initiative. Now, for 'New in Chess' interview referred to in the
example, the bishop on e4 is hanging and 28 Appreciation, for example, he insisted at
Ne6 is also a threat, hence Black's next. one point on showing his interviewer a
♘f6 28.♘e6 ♖c8 29.b3 A nice move, nice little King's Gambit miniature that he
defending the bishop on c4 and making room had won against an unnamed opponent, in
for the dark-squared bishop to switch to the a simultaneous display in Kiev 1941:
long diagonal. 3...♘f6 4.e5 ♘h5 5.♘c3 d6 6.♗c4 dxe5
♖h5 30.♘g5 ♗d5 31.♗d3+ ♔h6 32.♗b2 7.♘xe5 ♕h4+ 8.♔f1 .
Black is under constant harassment and still EXERCISE: Can you calculate a beautiful
15.♔c1? ♘f6 wins for Black, since the queenside, which cannot be said of the text
rook on e1 is undefended) 15...♘f2+ move. ]
16.♔d2 ♘e4+ with a repetition. Bronstein 17.♘f7+! ♔g8 18.♘d6 QUESTION: It
adds that the attempt to avoid the draw with seems that Bronstein, almost as a matter of
17.♖xe4 ♕xe4 18.♗d3? loses, but his principle, never moves an attacked piece.
variation is not quite correct. After ♕xg2+ ANSWER: Well, there was no doubt a sense
19.♔c3 , he gives ♗xh6? ( instead, Black of impish pride in th at, but it is also a
should clear the g7-square with tempo in a logical consequence of the position. When
d i f f e r e n t w a y : 19...♗xd4+! 20.♔xd4 you are material down, but with the initiative,
♘c6+ 21.♔c3 ♕g7+ when Black beats it is vital to keep threatening the opponent
off the attack and has a decisive material and not give him any respite. That often
advantage ) 20.♕xh6? ( but in this line, means ignoring attacks on one's own pieces
20.♗xh6! wins for White ) 20...♕g7 . ] and seeking ways to respond by creating
14.♕xe8 ♖xe8 QUESTION: But once the bigger threats oneself.
queens are off, surely White must be losing? ♖e6 19.♘e3 Now the knight finally moves,
He is a whole rook down, after all! but only after Black's rook has been lured to
ANSWER: Yes, but he still has a massive an exposed square on e6. The black
lead in development, and the black king is queenside pieces are never likely to see the
still not safe. light of day, and White's initiative has
15.♗d3!? become crushing.
[ Bronstein said it was impossible to ♗h6
calculate the consequences of the variation [ After 19...♗xd4 20.♘exf5 ♗xb2
15.♘xc7 ♖e4! , b u t a f t e r 16.♖e1! the obvious
White is clearly on top, as the computer A) but the computer finds the far-from-
shows: for example, ♖xf4 17.♖e8+ ♗f8 o b v i o u s , a n d e v e n s t r o n g e r 21.♖f1!
18.♗d3 ♔g7 19.♘xf5+ ♔f7 20.♖xc8 bringing the rook into the attack, and
and it is all over. ] threatening a deadly assault on the black
15...♘a6 16.♖xh1 QUESTION: What on king. For example, b5 (stopping Bc4)
earth is going on? 22.♘h6+ ♔g7 .
ANSWER: White has regained a piece, and is EXERCISE: Can you spot the lovely
now only an exchange down. He still has a forced mate?
large development advantage, with all his ANSWER: 23.♗e5+!! ♔xh6 ( or
pieces active and in threatening positions, 23...♗xe5 24.♖f7+ ♔xh6 25.♖xh7# )
while Black has his knight on h1 out of play, 24.♘f7+ ♔h5 25.♗e2+ ♔xh4 26.♖h1#
his queenside undeveloped, and faces a .
threat of Bxa6, followed by Nxc7. White has QUESTION: Beautiful! One does not
more than enough compensation for his expect to see such mating attacks once
material, as his attack continues. queens are off the board!
c6? ANSWER: Maybe not, but with so many
[ Obvious, but 16...d6 was probably the other pieces on, the position cannot really
only chance as it does at least make a be described as an endgame. It is more
small contribution to developing Black's a queenless middlegame, and with a
Nc3 is regarded as best). Instead, Tal being played in a 19th century spirit, a time
decides to get a piece out at last. ] when gambit pawns were accepted, almost as
4.♘f3 dxe4 5.♘xe5 ♕h4+?! a matter of honour.
[ In the spirit of 19th century fun, but ♗d4?!
objectively, 5...♘f6 is stronger, with equal [ 15...♗xe3 16.♗xe3 0-0 looks very strong,
chances. ] as after 17.0-0 ♗xd5 18.cxd5 ♖fe8
6.g3 ♕e7 7.d4 White's minor pieces are hanging
[ This disposes of the threats along the e-file, awkwardly on the e-file and the extra pawn
b u t i n f a c t , 7.♘c3 could also have been will soon fall. ]
played. Then the attempted larceny with 16.f5 ♗xd5 17.♘xd5 ♘xf5 QUESTION:
exd3 8.♗xd3 f6? fails to 9.♕h5+ ♔d8 This looks like a disaster for White! His king
( 9...g6 10.♗xg6+ hxg6 11.♕xh8 is caught in the centre.
is also clearly better for White) 10.0-0 ANSWER: Yes, but I think Tal was counting
with a ferocious attack for the piece. ] on the strength of the coming counterattack.
7...♘h6 8.♗c4 ♗e6 9.d5? This leads to 18.♕g4 ♘d6
great entertainment, but is objectively [ 18...♘ce7 was also good, but Bronstein
dubious. White is not sufficiently ahead in enters into th e spirit o f th e game, b y
development to justify this early opening of inviting the white queen into his parlour. ]
the position. 19.♘xd6+ ♖xd6 20.♕c8+ ♖d8 21.♘c7+
cxd5 10.♗xd5 f6 11.♘c4 ♔e7 22.♕xb7 I think this is probably what
[ 11.♕h5+ g6 12.♘xg6 hxg6 13.♕xg6+ Tal had seen when he played his 16th move.
♔d8 does not give White enough. ] QUESTION: Black's king looks very
11...♘c6 12.♘e3 Now White comes under ex posed! Surely it is now he who is in
pressure down the d-file, but it was difficult trouble?
to suggest anything better. ANSWER: It turns out that he is not, as
QUESTION: Why can't he just exchange on White has too few pieces with which to
e6? attack. In fact, it is the white king which
[ ANSWER: After 12.♗xe6 ♕xe6 13.♘e3 will soon come under a crushing attack.
♖d8 White's position is very unpleasant – ♕xc4 23.♗f4 Logically bringing up another
he is behind in development, his queen is piece.
exposed and he has also contracted some [ There is nothing to be had from the
nasty weaknesses on the kingside light discovered checks: for example, 23.♘d5+
squares. ] ♔f8 24.♘f4 ♗b6 and White is
12...♖d8 defenceless. ]
[ 12...0-0-0! looks even stronger. ] 23...♘e5
13.c4 ♕f7 14.♘c3 ♗c5 Offering a pawn to [ 23...♕c2! was even stronger, but the text
open more central lines. is good enough. Besides, we should not
15.♘xe4 QUESTION: That looks very criticize the play too severely. It is
risky! obvious that entertaining the crowd was
ANSWER: Yes, I am not sure Tal would the main purpose, rather than ruthless
have grabbed this pawn in a proper game. pursuit of the best moves. ]
But it is clear that the present encounter is 24.♖c1 QUESTION: Why doesn't he
(Diagram 30)
in view of what happens in the game, it take on g5, but now a fresh blow greets him.
seems that Black might as well try this, 15.♘g6! ♖e8 QUESTION: What happens if
although deep analysis shows that it still we take on g6 now?
loses. 14...hxg5 15.fxg5 and now Black [ ANSWER: Bronstein actually says this is
has a choice of evils: best, although after 15...fxg6 16.♕xe6+
A) 15...♘d5 16.♕h5 forces mate in ♔h8 17.♕xe7 Black is just losing: 18 Ne6
four,; is a terrible threat, and hxg5 ( 17...♕d8
B) w h i l s t a f t e r 15...♘xe5 16.dxe5 18.♘f7+ wins the queen) 18.fxg5 ♘h7
, Black also faces serious problems. The 19.♖xf8+ ♘hxf8 20.♕f7! leaves Black
only way to stay on the board for the defenceless against the threat of Bc4: ♗b7
time being is 21.♗c4 ♘h7 22.♗f4 ♕d8 23.♖e1
B1) 16...♘d5? again allows a forced and Re7 will decide. ]
mate after 17.♗h7+! with such lines as 16.♘xe6! QUESTION: This is amazing!
♔xh7 ( if 17...♔f8 18.♖xf7+! ♔xf7 [ ANSWER: Yes, it is delightful the way
19.♕h5+ ♔f8 20.♗g6 ♘f6 21.gxf6; Bronstein keeps sacrificing his knights,
or 17...♔h8 18.♕h5 and only various move after move. This is the third straight
desperate piece sacrifices on g5, f6 and move that he has offered one knight or the
f4 can delay the mate) 18.♕h5+ ♔g8 other, and now he self-forks them both!
19.♕xf7+ ♔h7 20.g6+ ♔h8 B u t n e i t h e r c a n b e c a p t u r e d : 16.♘xe6
21.♕xe8+; fxe6? ( while 16...fxg6? just leaves the
B2) 16...c4 17.♗c2 ♗c5+ 18.♔h1 queen en prise to 17.♘xc7 ) 17.♕xe6+
♘d7 , but here simply 19.♕e4 g6 ♔h7 18.♘xe7+ mates quickly. ]
20.♕xa8 wins, leaving White an 16...♕d6 Yet again, Black finds a way to
exchange ahead, and his queen is not avoid taking the knights, and yet again,
getting trapped, since the undefended White seems to have run up against a brick
rook on e8 means that Black cannot wall. How does he extricate his kamikaze
play ♗b7 .; cavalry?
C) QUESTION: So what should Black 17.f5!! "Banzai!". This stunning move
do? simply frees the f4-square for White's bishop,
ANSWER: The best (only) defence is to creating the threat of Bf4, followed by Nc7.
return the piece with 15...cxd4 16.cxd4 There is no defence.
♗b7 17.gxf6 ♗xf6 and hope to defend, QUESTION: Presumably Bronstein had to
but now White again wins with the have seen this when entering the whole
t h e m a t i c s a c r i f i c e 18.♗h7+! ♔f8 sequence starting with 14 Ng5, didn't he?
(relatively best, as other moves run into ANSWER: Yes, he must have done.
19 Qh5) 19.♗f4 and Black is just getting fxg6
mauled, without even having any extra [ There is nothing better: 17...fxe6 18.♗f4
material to show for his pains.; ♕c6 19.♗b5 ♕b7 20.♕xe6+ ♔h7
D) 15...-- ] 21.♗c6 wins everything. ]
So, taking on g5 is not possible. After the 18.♗f4 ♕c6 19.♗b5 ♕e4
text, Black looks again to have repulsed the [ 19...♕b7 20.♘c7 wins, ]
threats. With f7 defended, he threatens to [ whilst the desperado deflection sacrifice
Be3 and long castling gives White the pawn, in a thoroughly unclear position.
reasonable chances. ] The dark squares in Black's camp are very
7.♘g3 Revealing the point of White's 5th weak, but so is the e5-pawn. ]
move. The e4-pawn is defended. 14.♗h6 ♘d7 15.♗d3 f6? This turns out to
♘f6 8.e5 ♘e4!? be a serious mistake.
[ 8...♘d5 is more solid, when Black should [ Black should have accepted an endgame,
not have any problems. Instead, the text an exchange down but with two pawns,
instigates a long and sharp variation. ] after 15...♕xe5 16.♕xe5 ♘xe5 17.♗xf8
9.♕d3 ♘xg3 10.♕xg3 ♕xd4 QUESTION: ♔xf8 . Chances would be about equal. The
Is this right? Exchanging queens looks safer. text exposes the black king far too much. ]
ANSWER: Having said 'A' in the form of 8... 16.♖ae1!?
Ne4, I think Black should say 'B'. [ T h i s i s v e r y s t r o n g , b u t 16.♕xh7!
[ After 10...♕xg3+ 11.hxg3 White has looks even better, as 17 Bg6+ is a terrible
some advantage in the ending. The black threat. For example, ♕d4+ 17.♔g3!
light-squared bishop is lamentably passive ♕xe5+ 18.♗f4 ♕c3 19.♖ae1
and White has far more space. and the black king is dreadfully vulnerable,
QUESTION: But what about his doubled whilst she has no obvious way to complete
pawns? her development. ]
ANSWER: I don't think that is a 16...b6
significant factor. The pawns are not on [ 16...fxe5+ 17.♔g3 does not help at all.
an open file, and so are not really weak, White will continue with Qxh7 and Bg6+,
whilst the open h-file can be useful for with a lethal attack. Hunt's move attempts
White's rook. ] to develop the queenside, in the hope of
11.♕xg7 ♗xc3+ 12.bxc3 ♕xc3+ 13.♔f2 being able to evacuate the black king by
QUESTION: Crikey! Not exactly the sort of means of long castling, but she never gets
position one associates with the Caro-Kann, the chance. ]
is it? 17.♕xh7 ♕d4+
ANSWER: Indeed not, and that was probably [ The computer shows the line 17...♘xe5
part of the attraction for Bronstein. 18.♖xe5 fxe5+ 19.♗xf8 ♕d2+ 20.♔f3
QUESTION: What on earth is going on? ♔xf8 and now simply 21.♖f1 and Black is
ANSWER: Disappointingly, the position is defenceless against a discovered check. ]
probably just equal, if one believes the 18.♔g3 ♘xe5 19.♖xe5!?
computer. There is a distinct chance that [ Bronstein could not resist this sacrifice,
some sort of perpetual will result. However, which is still very good for White, but
any misstep by either side can be objectively, he should play the simple
immediately fatal. 19.♗g6+ ♘xg6 ( if 19...♔d8 20.♖d1 )
♖f8 QUESTION: Why not take a pawn with 20.♕xg6+ ♔e7 21.♗xf8+ ♔xf8 22.♖hf1
check? , winning.
[ ANSWER: There seems no objective QUESTION: Surely Bronstein could see
reason not to. There could follow that, couldn't he?
13...♕xc2+ 14.♗e2 ♕c5+ 15.♗e3 ♕f8 ANSWER: Yes, of course, but he simply
16.♕g4 and White has compensation for could not resist the more spectacular line.
He had doubtless seen his 21st move, and 25.♗xf7 ♕xf7 26.♗d6+ ♔f6
once seen, such a move is hard to refrain [ Or 26...♔e8 27.♕h8+ ♔d7 28.♖d1
from, especially for such a romantic as and there is no defence: for example, ♗b7
Bronstein. ] 29.♗b8+ ♔e7 30.♕d8# . ]
19...♕xe5+ 20.♗f4 ♕c3 White needs to 27.♕h6+ ♕g6 28.♗e7+
unpin his bishop, after which the attack will [ 28.♗e7+ ♔f7 29.♕f8# is mate, and
be crushing. Therefore... otherwise the queen is lost.
21.♔h4!! It is disappointing that this should QUESTION: Presumably, Bronstein would
be the computer's first choice, after just a have been choked to discover the missed
mi l l i s e c o n d o f t h o u g h t . B l a c k i s q u i t e opportunity at move 24?
defenceless. ANSWER: Actually, I am not sure he
f5 Shutting out the bishop from g6, but now would. Yes, it spoils the integrity of the
it simply reroutes to another diagonal. finish, but the remarkable defensive
22.♗e2! ♕f6+ 23.♔h3 Steinitz would resource 27...Rh8+ 28 Kg3 Rh7, trapping
doubtless have loved the way in which the the bishop on f7, would have delighted
white king is able to defend itself in this way. Bronstein, I think. He was very much an
♖f7 artist, who loved the richness of the
[ QUESTION: But wait a minute, isn't resources on the chessboard, so any
23...♖h8 trapping the queen? disappointment at the flaw in his game
ANSWER: No! White continues simply would have been comp ensated for by
24.♗h5+ unpinning, and forcing mate in seeing such a lovely defensive trick. ]
two: ♕g6 25.♗xg6+ ♔d8 26.♗c7# . ] 1-0
24.♗h5? A blunder, which could have cost
Bronstein the fruits of his play.
[ There was a simple win with 24.♕g8+ 21 C13
♖f8 25.♗h5+ ♔e7 26.♕h7+ ♖f7 27.♗xf7 Kaplan,J
♕xf7 28.♗d6+ ♔e8 29.♕h8+ ♔d7 Bronstein,D
30.♖d1 , etc. ] Hastings (Game 12) 1975
24...♔e7? Returning the compliment. [Steve Giddins]
[ After 24...♗d7! 25.♖d1 0-0-0 26.♕xf7
♕xf7 27.♗xf7 it looks as though White is 1.e4 e6 2.d4 d5 3.♘c3 ♘f6 4.♗g5 dxe4
winning easily, with his extra piece, but 5.♘xe4 ♗e7 6.♗xf6 gxf6 QUESTION:
the eagle-eyed computer points out that the That is a surprise? Can this really be good?
bishop on f7 is not escaping after ♖h8+ ANSWER: It looks odd at first sight, but it is
28.♔g3 ♖h7! . The position is only a draw a reasonable option that has been known
after 29.♗xe6 ♗xe6 . about for a century or more. By recapturing
QUESTION: Amazing! with the pawn, Black strengthens his central
ANSWER: Yes, computers are desperately control and also has the possibility of using
tricky in such positions. I am sure neither the half-open g-file. It has obvious
player even suspected the existence of such similarities with the Caro-Kann line 1 e4 c6
a fiendish trick, and one cannot blame 2 d4 d5 3 Nc3 dxe4 4 Nxe4 Nf6 5 Nxf6+
them. ] gxf6, which was also a favourite of
Bronstein's. position. ]
7.♘f3 ♘d7!? This was a favourite move of 8...c5 Annotating the game in the January
Bronstein's in this position. 1976 issue of 'CHESS', Bronstein confessed
[ The more common treatment is 7...b6 that he was slightly nervous about playing
, by means of which Black aims for a this move, but felt that "it had to be done".
restrained development, along the lines of 9.d5
...Bb7, ...c6, ...Qc7, ...Nd7 and ...0-0-0. [ White should probably have played
This somewhat Hedgehog-like set-up is 9.0-0-0 when Bronstein intended f5
very solid and offers Black reasonable 10.♘c3 cxd4 11.♕xd4 ♗f6 followed by
chances. It has been played a number of castling short.
times by Ulf Andersson. ] QUESTION: Again, this looks pretty shaky
[ The line with 5...gxf6 experienced a to me. Isn't short castling dangerous, with
revival of popularity about 20 years ago, the g-file open?
when the restlessly creative Alexander ANSWER: Even more to the point, Black
Morozevich essayed it a number of times, cannot get his king castled in this line after
with a different interpretation, starting with 12.♕a4! , -- ( since 12...0-0? loses a piece
7...a6 . In this treatment, Black plays an after 13.♗b5 .; It seems to me that Black is
early ...b5, with an extended fianchetto. i n s o m e t r o u b l e h e r e , a s e v e n 12...a6
Morozevich won several nice games with does not prevent 13.♗b5 , when White's
the line. It is no coincidence that he and position still looks threatening after ♖b8
Bronstein should both be attracted to 5... 14.♗xd7+ ♗xd7 15.♕f4 ♖c8 16.♖he1
gxf6, as they were similarly creative in , etc; ♗xc3 17.bxc3 ♕c7 18.♕d4
their approach to the game. ] does not solve Black's problems. )]
8.♕d2 9...f5 10.dxe6? Bronstein records that this
[ An old Bronstein experience with the same was played after long thought, but it seems
line saw his opponent try to break up the to be a clear mistake, after which Black
b l a c k p a w n f r o n t w i t h t h e i mmed i a t e escapes from his problems and even assumes
advance 8.d5 . However, this failed to the better of things.
yield anything, and after the sequence ♘b6 [ Instead, 10.♘c3 keeps up the pressure,
9.♗b5+ ♗d7 10.♗xd7+ ♕xd7 11.dxe6 when Black's position is still rather
♕xe6 12.♕e2 0-0-0 13.0-0 f5 14.♘g3 unpleasant.
♕xe2 15.♘xe2 ♗f6 Bronstein later QUESTION: It doesn't look as though
claimed Black is already slightly better in Bronstein's opening was up to much! Does
the endgame, and he went on to win in this mean that 5...gxf6 is no good?
Averbakh-Bronstein, USSR Championship ANSWER: Not necessarily. I think
Semi-final, Leningrad 1946. Bronstein's follow-up with 7...Nd7 is
QUESTION: But why should Black be probably inaccurate, fond though he
better here? His pawns are doubled. obviously was of the move. In general,
ANSWER: Yes, but they are not on open Black has to be quite careful in the early
files, so their vulnerability is limited, stages of these positions, as if he breaks
whilst Black's bishop is the strongest out prematurely with moves such as ..c5
minor piece on the board, in this open and ...f5, his lack of development and
uncastled king can rebound on him. I the long diagonal, but White has no other
don't think Black has any serious problems pieces that can support the attack, whilst
in the Andersson line with 7...b6, but such Black's dark-squared bishop does an
a rather restrained and passive set-up is excellent job of protecting the dark squares
not to everybody's taste. ] around his king.
10...fxe4 11.exd7+ 13.♘d2 ♕f5 14.0-0-0 ♕xf2 15.♘xe4?
[ 11.exf7+ was the alternative. Bronstein Bronstein identifies this as the decisive
reports that his opponent rejected it because mistake As he puts it, "White ties himself up
of ♔xf7 12.♕f4+ ♘f6 13.♘e5+ ♔e8 regaining the pawn".
14.♗c4 ♖f8 . However, Bronstein did not [ He recommends 15.♗c4 (at least, I think
like this at all. that is what he recommends – 'CHESS',
QUESTION: So what do you think? ] which was in descriptive notation in those
[ ANSWER: Well, Bronstein's instinct was days, actually gives the illegal move 15 B-
correct, even if not his variations. In the N4) "with equality", an assessment with
first place, 11.exf7+ ♔xf7 12.♕f4+ ♘f6 which the computer agrees. ]
13.♘e5+ ♔e8 does in fact lose by force – 15...♕f4+ Forcing the knight back into a self-
i n s t e a d o f 14.♗c4? , ( White just plays pin. From now on, Black's pieces flow into
14.♗b5+ and Black can resign at once. )] the game with threats at almost every move.
[ After 11.exf7+ ♔xf7 12.♕f4+ 16.♘d2 ♗g4 17.♖e1 ♗g5 The pin is deadly,
Bronstein's intended ♔g7 is therefore the as White is unable to evacuate his king to b1,
correct move, when he thought he was whilst Black threatens to bring a rook to the
allowing perpetual check, but the computer d-file.
points out that this is not so. Black is just 18.♗d3 ♖ae8 19.♖ef1 QUESTION: Does
much better after 13.♕g4+ ♔f8 14.♕f4+ White have anything better?
♗f6 and the checks run out. ] [ ANSWER: Sadly not. For example,
11...♕xd7 12.♕c3 This is the critical 19.♖xe8 ♖xe8 20.h3 and now simply ♗c8!
position. The other reason Kaplan did not followed by ...Qf2, the point being that
take on f7 last move is that he assessed this 21.♖f1? allows mate in two after ♕xf1+
position as being good for him. 22.♗xf1 ♖e1# .
QUESTION: Well, surely it is, isn't it? QUESTION: It is shocking how quickly
ANSWER: It looks that way, as the black the position has turned against White. How
rook on h8 is hanging, White is ahead in did this happen?
development, and he has the better pawn ANSWER: It is all the consequence of his
structure. However, things are not so simple, 15th move, which landed him in a fatal
as Black proves able to untangle successfully, pin. ]
after which his two bishops prove very 19...♕e3 20.h3 ♗e2 21.♖f5
strong. [ 21.♗xe2 ♕xc3 22.bxc3 ♖xe2 23.♖d1
0-0 QUESTION: Is the king really safe here? ♖d8 nets a piece at once. ]
It looks exposed. [ Q U E S T I O N : B u t w h a t a b o u t 21.♖e1
ANSWER: The point is that White does not , pinning Black's own bishop?
have the pieces with which to harass the A N S W E R : T h e n t h e r e f o l l o w s ♕f2
enemy king. The queen looks very strong on when White is still tied hand and foot. The
8...exd5 9.cxd5 QUESTION: This position shown that this is not best.
looks like absolute suicide for Black! Surely [ Kasparov quotes 11.♗e2 as giving White
the central pawns are going to kill him? a clear advantage, but none of this was
ANSWER: Well, Alekhine's Defence is one known at the time of the present game. ]
of the most archetypal of hypermodern 11...♗xf3 12.gxf3 ♗b4 13.♗xc4 0-0
openings, and the whole idea is to encourage QUESTION: So what is going on here? The
White to set up a broad pawn centre, in the position looks a real mess already.
hope that it will prove too unwieldy and will ANSWER: Believe me, this position is a
serve as a target for Black's guerrilla raids picture of calm and order, compared with
from the edge. The 6...c5 variation is an what will follow. Both sides are playing
extreme application of this approach and logically. Black has to get developed and
certainly not for the faint-hearted. start creating threats against the white centre.
c4 Already, in defending his two central passed
[ This was a relatively new idea at the time pawns, White has allowed the capture on f3,
of this game, replacing the discredited splitting his pawns and exposing his king,
older alternative 9...♕h4+ 10.g3 ♕d4 and it is still far from clear where the latter
, after which 11.♗b5+ ♗d7 12.♕e2 ♘xd5 will find a refuge – castling on either side
13.e6! fxe6 14.♕xe6+ ♘e7 15.♘f3 ♕f6 will leave it exposed, and in the centre it is
16.♕e2 was found to be good for White. even more insecure. Black's immediate threat
Indeed, one of the games referred to by is 14...Nc6 15 Qe4 Nxe5, demolishing the
Kasparov, which reached this very position, white centre.
was Ljubojevic-Noses, Dresden 1969. As 14.♖g1
we will see, the entire 6...c5 line became [ White sets up threats along the g-file, and
a favourite battleground for the young with 14.♖g1 also prepares to meet ♘c6?
and highly talented Yugoslav GM. ] with 15.♕g4 , winning. ]
QUESTION: So what is the point of 9...c4? 14...g6!
ANSWER: The move does several things. It [ Ljubojevic's prior experience with this
shuts in White's king's bishop, and it line had also encompassed the white side
prepares to bring its counterpart out, either of this position. His opponent had played
to c5, cutting through the heart of White's 14...♕c7? and been crushed after 15.e6 f6
position, or to b4, pinning the knight and ( or 15...fxe6 16.♗h6 e5 17.♗xg7! exd4
thus exerting pressure against the d5-pawn. 18.♗xd4+ ♔f7 19.♖g7+ and wins –
The whole of Black's play is about targeting Kasparov ) 16.♗h6 ♕xc4 17.♖xg7+ ♔h8
the two white pawns on d5 and e5 – if White 18.♖g8+! ♔xg8 19.♕g1+! and Black
manages to maintain them and get castled, resigned in Ljubojevic-Honfi, Cacak
then he will just have a crushing advantage, 1970. ]
so Black must fight tooth and nail against 15.♗g5! Even this was all theory at the time.
this. [ A correspondence game between two
10.♘f3 ♗g4? Objectively, this has to be British amateurs, Gibbs and Stewart, had
condemned. seen White try 15.♗h6? , but he ran into
[ Modern theory prefers 10...♗b4 . ] ♘c6 16.♕e4 ♘xe5! and was massacred.
11.♕d4 Once again, the passage of time has Bronstein's move was a key improvement,
these complications, is that Black might as declined. We can only be grateful that he did
well take the rook? so, as well as marvelling at Bronstein's
ANSWER: Yes. conduct of the rest of the game, with so little
18.d6 time left.
[ Ljubo's experiences with this line are 18...♕c8? This proves to be a serious
highly interesting. As we have seen, prior mistake.
to the present game, he had played the [ 18...♕c6? is no better, losing at once after
white side a couple of times, and always 19.e6 fxe6 20.♗xe6+ ♔g7 21.♗h6+
won. Then he switched to the black cause and mate. ]
for the present game, and lost. However, [ After 18...♕c6? 19.e6 , ♘8d7
undeterred, he carried on playing it as is actually the best try, but also good for
Black, and racked up several victories. One White after 20.exf7+! ♔g7 21.0-0-0 .
of these came later the same year, when the QUESTION: How come a position like this
Romanian GM Florin Gheorghiu took him is good for White? He is still a rook down,
on, and attempted to improve on and what is he threatening? Say, I play
Bronstein's play with the move 18.♔e2? simply ♗c5 , getting the bishop back to
. However, this proved bad, and Gheorghiu safety?
lost as follows: ♕c5! (as we will see below, ANSWER: White still has a crushing attack,
this is a key resource for Black in many and can take his time. Simply 22.♔b1!
variations; the black queen defends the gets the king off the c-file and prepares 23
bishop on g1 and, together, the two pieces Ne4 . The co mp uter sho ws th at Bl ack
create threats against the white king) c a n n o t d e f e n d : f o r e x a m p l e , ♖ac8
19.♖xg1 ♕xg1 20.♗f6 (it looks as though (pointless, but Black has no way to get any
Black will be mated after Qh6, but the pieces over to the kingside or to shut out
white king is too exposed) ♕g2+ 21.♔e3 W h i t e ' s a t t a c k i n g u n i t s ) 23.♘e4
( 21.♔e1 ♘8d7 is no better: for instance, and there is no defence to 24 Bf6+.
22.♕h6 ♘xf6 23.exf6 ♖ae8+ 24.♘e4 QUESTION: It seems amazing that with so
♖xe4+ 25.fxe4 ♕xe4+ 26.♔d2 ♕d4+ many pieces on the board, and an extra
and the f6-pawn drops) 21...♕xb2 22.♔d3 rook, Black cannot defend.
( or 22.♕h6 ♕c1+ ) 22...♘8d7 23.♘e4 ANSWER: The problem is that his extra
♖ac8 24.♕h6 ♘xe5+ 25.♔e3 ♖c3+ rook is not really taking an active part in
0-1, F.Gheorghiu-L.Ljubojevic, Manila the game, and the d6-pawn cuts his position
1973. ] in half. Meanwhile, every white piece is
Kasparov's 'My Great Predecessors' volume taking part in the attack, apart from his king.
reveals one other stunning thing about this If you have trouble believing that White
position. He claims that Ljubo recently is winning here, it is worth spending some
revealed to him that, after playing 18 d6, time, pushing the pieces around and trying
Bronstein offered a draw, because he had to find a defence. You will learn a great
just five minutes (!) left on his clock. Ljubo, deal about the power of the initiative. ]
who, as we have seen, was armed to the teeth [ QUESTION: So if the other queen moves
in this opening line, had used just ten lose, what should Black do?
minutes on his own clock, so naturally he ANSWER: The critical move is 18...♕c5!
follow-up with g3 and play a kind of Closed controls d4, and can later come to e3. ]
Sicilian structure. Paul Keres was one top 9...0-0 10.g4 As predicted, the moment
player who adopted it a few times. Black castles and takes his rook off the h-file,
d6 3.g3 ♗g4 Bronstein seizes the chance to advance g4,
[ 3...d5! , as pointed out by Fischer in 'My setting his kingside pawns in motion and
60 Memorable Games', is quite effective preparing to castle.
here, despite the tempo loss, since 4.exd5?! ♗c6!? QUESTION: What is this about?
♕xd5 is rather awkward for White. ] ANSWER: I guess Black's logic was that,
4.♗g2 ♕c8 5.h3 ♗d7 QUESTION: I am now the bishop no longer has anything to do
surprised by 5 h3. Now White is unable to on the c8-h3 diagonal, it redeploys to the
castle, as his h-pawn is attacked. long diagonal, to oppose the white king's
ANSWER: True, but Bronstein intends to bishop.
deal with this by a later g4. In such closed QUESTION: But I sense you disapprove?
positions, White often advances f4 and g4 ANSWER: Yes, I am not entirely convinced,
anyway, so providing he waits until Black mainly because it leaves the queen's knight
has castled (thus reducing the force of the without a decent square.
reply ...h5), he should be fine; indeed, he [ The move 10...♘c6 looks more natural,
may well gain a tempo over the immediate ... although the black queen is not ideally
Bd7. placed on c8 either, and will probably have
6.d3 g6 7.♗e3 ♗g7 8.c3 QUESTION: I to lose a tempo to come to c7 at some
thought in the Closed Sicilian, the knight point. ]
usually came to this square? 11.0-0 QUESTION: So how do you assess
ANSWER: In the main Closed Sicilian, that the position?
is indeed the case, but the knight is not ANSWER: I prefer White's game, although it
always so well-placed there, since it later is perhaps not a lot. He has a typical
gets hit by ...b5-b4, and Black can utilize the development for these Big Clamp systems,
d4-square for his knight. White often ends with his pawns on c3, d3 and e4 controlling
up playing contortions such as Qd2, Nd1 and lots of central squares, and he has already
then c3, to evict such a knight. Having the started advancing his kingside pawns, which
option of putting a pawn on c3 is one of the is his main plan here. Black, by contrast, is a
points of this 2 Ne2 set-up. little slow. He has not yet completed his
Positions with this arrangement were development, and his pieces stand slightly
explored in considerable detail by the awkwardly, whilst his customary queenside
inventive Canadian IM, Lawrence Day, who counterplay with ...b5-b4 has not yet got
christened the set-up 'the Big Clamp' system, going.
and even published a pamphlet on it. British ♘e8?! I do not especially like this either,
GM Nigel Davies is another who has although it is true that it will soon be forced
extensive experience with such variations. after an eventual g5. I should prefer to get on
♘f6 9.♘d2 with the queenside development and start
[ 9.♘a3 followed by Nc2 is another plan, some counterplay on that flank.
and the one generally preferred by 12.♘g3 ♘d7 13.f4
Lawrence Day. From c2, the knight [ QUES TION: W a s n' t 13.d4 possible
a clear advantage for Black, but I am not often the key to understanding how an
convinced this assessment can be entirely attack can s ucceed . I t is n ot the to tal
trusted, for the reasons outlined above. I material balance on the board as a whole
think White has compensation here. ] that matters, but the balance in the key
20.f5 Now even the computer thinks White sector, where the action is taking place. ]
has sufficient compensation to balance out 22.fxg6 hxg6 23.♘g5! ♘e8?
the material, which I think bears out my [ The knight obviously cannot be taken, in
previous comment. view of forced mate: 23...fxg5? 24.♖xf7
QUESTION: How do you mean? ♔xf7 25.♕g7+ ♔e8 26.♕f8# . ]
ANSWER: Well, the only difference in the [ However, although Black must lose back
position from the variation after 19...Nc4 20 the exchange, he does have a major
Bc1 Bd7 21 f5 is that the white bishop is on decision to make, in terms of which white
c1 instead of e3, and the black knight on c4 minor piece he should get for the rook.
instead of b6. I agree that these changes are Ratner's choice was the knight on g5, but
slightly in Black's favour, since his knight is the computer heartily condemns this and
more actively placed, but I really don't see advocates 23...♖g7 instead, choosing to
that the difference is that great. eliminate the white queen's bishop.
f6 21 Bh6 was threatened, which would QUESTION: What do you think?
regain the exchange and force this ANSWER: It is not an easy decision over
weakening move anyway, so Black had little the board, but the variations seem to
choice. s u p p o r t t h e c o mp u te r 's c h o i c e . A f t e r
21.♗h6 ♖f7 This results in White regaining 24.♗xg7 (not obligatory, but I am not sure
the exchange, thanks to a tactic, but Black White can gain anything much from
did not have a lot of choice. d e f e r r i n g t h e m o v e ) ♔xg7 25.♘e2
[ If 21...♖e8! he leaves the f6-square too White retains the initiative, but Black is
vulnerable, and after 22.g5! the defensive probably OK. It seems this was his only
barriers around the black king prove to hope, as the game continuation sees his
have all the strength and resistant position collapse quickly. ]
qualities of damp polystyrene. 24.♘xf7 ♔xf7 25.e5! This is the real killer
The other important point to note about follow-up, which Black probably missed or
this position is the way the central pawn underestimated. The pressure against f6 and
configuration, especially the pawn on e7, the other dark squares around the black king
cuts Black's position in half, and prevents is too great to resist.
his pieces on the queenside from getting dxe5 26.♘e4?!
across to help in the defence of their king. [ A creative follow-up, but the computer's
Black's queen, rook, bishop and two straightforward 26.♕xe5 was probably
knights are all pretty well sealed off from stronger. Now Black gets another chance to
the kingside, whilst White is attacking on hold his position together. ]
that wing with almost his entire army. 26...♘c4 27.♘g5+ ♔g8 28.♕b1
QUESTION: We have seen this scenario QUESTION: What is going on now?
before in this book! ANSWER: This last move was the point of
ANSWER: We certainly have, and it is so Bronstein's idea. He attacks g6, which
cannot be defended other than by the advance 33.♕g3! Calmly transferring the queen to h4,
of the f-pawn, which obviously weakens where the threats will be decisive.
Black's position enormously. However, it e6 34.♘xe6 ♕e7
seems he can still defend. [ 34...♗xe6 35.♗xe6 leaves Black unable
f5 29.gxf5 gxf5 to cover g8, without unprotecting g7. ]
[ 29...♗xf5! was better, activating the black 35.♘f8 Black resigned, as again he cannot
pieces. Then after 30.♕e1 ♘f6 31.♕g3 protect the g8-square and avoid 36 Ng6+ as
♘h5 Black can hope to repulse the well. The finish is a lovely early example of
threats. ] Bronstein's attacking prowess, but what
30.♕d3 ♕d8? This proves fatal, by allowing lends the game special interest is the striking
White's next move. exchange sacrifice on move 18, with its eerie
[ Black had to blockade the d5-pawn with presaging of Bronstein's later famous
30...♘ed6 , when the complicated variation Grünfeld innovation.
31.♘e6 ♗xe6 32.♕g3+ ♔f7 33.♕g7+ 1-0
♔e8 34.dxe6 ♕xe6 35.♕f8+ ♔d7
36.♕xb8 ♕xh6 ends in perpetual check
after 37.♖xf5! ♘xf5 38.♕xb5+ ♔d8 24 B83
39.♕b8+ ♔d7 40.♕b5+ etc, since if ♔e6 Bronstein,D
41.♕xc4+ ♔f6 42.♕a6+ e6 43.♕xa7 Levenfish,G
♕c1+ 44.♔h2 ♕f4+ it is Black who gives USSR Championship, Moscow (Game 15)
perpetual. ] [Steve Giddins]
31.d6! Clearing the path for the light-squared
bishop to enter the game, with decisive effect. 1.e4 c5 2.♘f3 ♘c6 3.d4 cxd4 4.♘xd4
It is a nice feature of the game that the ♘f6 5.♘c3 d6 6.♗e2 At the time of this
attack initially concentrated on the dark game, this was the main move here, but in
squares around the black king, but then view of the strength of Black's reply, it fell
moves such as 28 Qb1 and now 31 d6 switch out of fashion.
it to the light squares. [ For a long time, White used to play the
♘cxd6 Sozin line 6.♗c4 , ]
[ Q U E S T I O N : W h a t h a p p e n s i f 31...e6 [ but that in turn has now given way
to keep the light-squared bishop out? overwhelmingly to Rauzer's 6.♗g5 . ]
ANSWER: Then the computer shows a 6...e5 The Boleslavsky System, the
b r i l l i a n t w i n b y 32.♖xf5! , crashing granddaddy of all these ...e5 lines in the
through on those same light squares. For Sicilian.
example, exf5 33.♗d5+ ♔h8 34.♘f7+ QUESTION: It was quite a revelation in its
♔h7 35.♘xd8 ♖xd8 and now simply day, wasn't it?
36.♗g5 ♖c8 37.♕g3 finishes Black off in ANSWER: Yes, it revolutionized our
short order (it is mate in four, in fact). ] understanding of such structures.
32.♗d5+ ♔h8 Boleslavsky (whose daughter eventually
[ 32...e6 33.♖xf5! is another brilliant married Bronstein) showed that Black's
finish: ♘xf5 34.♕xf5 ♕e7 35.♕g6+ dynamic play compensates for the apparent
♔h8 36.♘f7+ , etc. ] weakness of the d5- and d6-squares.
7.♘b3 ♗e7 8.0-0 ♗e6 9.f4 exf4 10.♗xf4 probably only equal, although Boleslavsky
0-0 QUESTION: How would you assess the gives the impression he thought White
position? was better here. ]
ANSWER: Chances are equal. White has 14.exf6 ♘xf6!?
more space, because of his central pawn on [ Boleslavsky rejects 14...♗xf6 because of
the fourth rank, and he still potentially has 15.♘xe4 dxe4 16.♘c5 "winning a pawn",
an outpost on d5, and a weakness to attack but the computer continues the variation
on d6. But he also has a weak pawn on e4, with exd3 17.♘xe6 dxc2! 18.♕xc2
and Black has the half-open c-file. Black's ( g r a b b i n g t h e e x c h a n g e b y 18.♕xd8
king is also a little more secure, White's ♖fxd8 19.♘xd8 ♖xd8 is actually bad for
h a v i n g b e e n e x p o s e d s o me w h a t b y t h e White, because the c2-pawn is too strong:
advance f4. for example, 20.♖ac1 ♘b4 21.a3 ♘d3
11.♔h1 d5?! QUESTION: Why is this 22.♖xc2 ♘xf4 wins material, due to the
wrong? Surely, ...d5 is the advance Black back-rank weakness) 18...♘d4 19.♘xd4
usually wants to play in the Sicilian, and it ♕xd4 , with dead equality. ]
gets rid of his weak pawn. 15.♕e1 ♕d7 16.♘a4 ♗f7? Now Black gets
ANSWER: All of that is true, and at first into trouble.
sight, one would expect Black to be doing [ The threat of 17 Nac5 looks unpleasant,
fine if he can get in ...d5. But Boleslavsky but typical Sicilian dynamics come to his
himself, who annotates the game in the aid: 16...♘e4! with the point that 17.♗xe4
Russian tournament book, criticizes the move dxe4 18.♕xe4 ♗d5 19.♕e2 ♖ae8
as "not in the spirit of the position" and gives Black adequate compensation for the
allowing White to obtain a positional pawn, in the shape of his two strong
advantage. bishops and the exposed white queen. ]
[ I n s t e a d , h e r e c o m m e n d s 11...♕b6 17.♘ac5 ♕c8 18.♗g3
, in t en d in g . .. N e5 , a n d th e n i f 12.♘d5 [ 18.♗g5! was stronger, as pointed out by
♗xd5 13.exd5 ♘e5 14.c4 ♘e4 Boleslavsky, with ideas of Bf5 and also
, as had all been played two years earlier in Qh4, attacking the kingside. ]
a game A.Sokolsky-D.Bronstein, Leningrad 18...♗xc5!?
1947. ] [ Surrendering the bishop-pair is a long-
12.e5 ♘e4 term concession, but I am not convinced
[ Boleslavsky mentions 12...♘d7 13.♘xd5 that the computer's malodorous suggestion
♘dxe5 14.c4 , and assesses the position as of 18...♗g6 is really any better. After
highly favourable for White. That seems 19.♗xg6 hxg6 20.♖d1 the weak black
an exaggeration, but White has an edge, pawns should guarantee White a lasting
thanks to his queenside majority. ] edge. ]
13.♗d3! f5 QUESTION: I don't understand 19.♘xc5 ♖e8 20.♕f2 QUESTION: Things
this. Why not double White's c-pawns? seem to have gone a little wrong for Black?
[ A N S W E R : A f t e r 13...♘xc3 14.bxc3 ANSWER: Yes, now White has a nice, and
the pawns are less weak than they look, long-lasting positional advantage. He has
and meanwhile, White has a nice outpost two bishops, and excellent control of the
on d4 for his knight. Even so, chances are weakened dark squares in Black's position.
♘e5 21.♗f5 ♕c6 22.♖ae1 ♘eg4 rook endings (it appeared under the joint
23.♕d4!? Probably not the most accurate, names of himself and Smyslov, but the
although White is still better. latter's name was only added in order to get
[ Boleslavsky says 23.♖xe8+ ♕xe8 24.♕f4 the book in print; Levenfish had always been
♘h6 i s w e a k e r , b u t a f t e r 25.♗d3 viewed with suspicion by the Soviet
White keeps a solid advantage, authorities, as he had a pre-Revolutionary,
( whereas Boleslavsky only considers the bourgeois background).
weaker 25.♘xb7 . )] QUESTION: So, all the more reason why he
23...b6?! QUESTION: You don't like this? should have known better!
ANSWER: I think this time Boleslavsky was ANSWER: Yes, but paradoxically, this may
right on the mark, in saying that the text have been precisely why he entered this
leads to the exchange of queens and a ending. I suspect Levenfish thought that a
difficult endgame for Black. young master such as Bronstein, who was
[ His recommendation is also the computer's known for his fiery imagination and tactical
first choice: 23...♘h6 24.♗d3 b6 25.♘b3 play, would find it much harder to outplay
♘e4 "with some counterplay for Black". such an experienced endgame expert as
W h i t e r e m a i n s b e t t e r a f t e r 26.♗f4 himself in a position like this, whatever the
, but a position with queens on offers objective merits of White's position. But if
Black more chances of creating counterplay so, he proved very wrong, as Bronstein gives
than a lifeless technical ending. ] a textbook demonstration of how to handle
24.♘d7 ♕c4 such positions.
[ 24...♘xd7 25.♕xg4 ♘f6 26.♕h3 ♘e4 28.♔g1 ♗g6 29.♗d7 QUESTION: He
is also unpleasant, but would have the turns down the chance to double Black's
same merit as the line in the previous note, pawns?
of keeping queens on. ] ANSWER: Yes, there would be no value in
25.♘xf6+ ♘xf6 26.♗e5 ♕xd4 27.♗xd4 doing that, if it costs White his bishop-pair.
QUESTION: So White has the two bishops QUESTION: But he would still have bishop
in the endgame. versus knight.
ANSWER: Yes, and this position is solid ANSWER: Yes, but that is usually
masochism for Black. As well as the two significantly less of an advantage than two
bishops, White also has one fewer pawn bishops against bishop and knight.
island. The computer may assess White's QUESTION: Why?
advantage as no greater than in the previous ANSWER: Because the bishop itself has one
couple of notes, where the queens stayed on great disadvantage against a knight, even in
the board, but in practice, I think the game an open position – it can only control
position is much harder and more unpleasant squares of one colour. So, if White were to
to defend as Black. exchange light-squared bishops here, his
QUESTION: So why did Levenfish play this remaining bishop on d4, splendid piece
way? though it is, can only control half the squares
ANSWER: That is a good question. He was on the board, whereas the black knight can
himself a great endgame specialist and a go to all 64. This 'colour-blindness' of a
decade or so later, he wrote a classic book on bishop is a major issue, which is often
[ 43...♘e4+ avoids either weakening pawn ♔e6 QUESTION: Can't he take on g4?
move for the moment, but the knight is [ ANSWER: He can, but after 50...♘xg4
unstable on e4 after 44.♔f3 . The threat of 51.cxd5 White has a passed pawn, which
c4 hangs in the air. ] will be a serious trump card. ]
44.♗c2 ♘c4 45.♗c1 QUESTION: It looks 51.♗b2 ♘f7 52.cxd5+ ♔xd5 QUESTION:
as though Black is fighting back? I don't understand what White has achieved
ANSWER: No, this is just an illusion. Yes, over the last few moves. Hasn't he just
for the moment, the white bishops appear exchanged off Black's weak d5-pawn for
passive, but the knight can be driven away him?
again with b3, whenever White wants, and ANSWER: Yes, but he has also opened more
meanwhile, nothing essential has changed. lines and cleared more space for his pieces.
♔d7 46.♔g3 ♔e7 47.♔f4 See the note to The d5-pawn was never in real danger of
White's 43rd move. Having provoked the dropping off, but it was depriving the white
move ...g6, weakening the dark squares on pieces of some squares and lines that would
the kingside, White attempts to infiltrate on have been useful. Now, for example, the a2-
those weak squares with his king. Black g8 diagonal is open, so Black must worry
must prevent Kg5. about the white bishop getting round the
♔f6 48.h4 Stopping Black playing ...g5+ back to g8.
himself, and asking him what he is going to 53.♗e4+ ♔d6 54.♗g7 QUESTION: Now
do next. what is he doing?
QUESTION: It all seems so slow, like White ANSWER: The bishop just controls some
is fighting for trifles! squares, preventing a possible ...Nh6,
ANSWER: That is the nature of endgame attacking g4, and also setting up a potential
technique. Slow, systematic probing, and the Bf8+, driving the black king back.
accumulation of tiny advantages is the order QUESTION: You say "a possible" this and
of the day. Eventually, the straws will mount "a potential" that. You don't seem sure if
up, until one of them breaks the camel's back. these are threats or not!
♗d7 49.b3 ♘e5!? ANSWER: No, I am not. That, again, is how
[ 49...♘d6 would avoid the tactic that endgames work. Bf8+ may never happen, but
follows, but it does not essentially change it might. Either way, it is yet another thing
anything. White can just continue with for Black to worry about. The possibility of
50.♗e3 , preparing a check on d4, and if Bf8+ may persuade him to play in a different
♘b5 then 51.a4 ♘xc3? 52.♗d4+ . ] way from otherwise, and White may then be
50.c4! QUESTION: A tactic at last! able to exploit that. It is all part of the
ANSWER: Indeed. Another aspect of good process of just grabbing every little
endgame technique is not to forget about advantage that one can – to repeat the earlier
tactics. The play may be slow and full of analogy, one never knows which straw will
p r o t r a c t e d ma n o e u v r i n g , b u t t h e g r e a t be the one that breaks the camel's back, so
endgame players never forget about tactical one just goes on gathering as many as
opportunities, and make full use of them possible, and waits to see.
when they appear. Here, the threat of 51 Bb2 ♗e6 55.b4 ♔e7 56.♗d3 ♘d6 57.h5
forces a concession from Black. Now that the knight has left f7, there is no
longer the move ...g5+ for Black, so White the white king: ♔e6 63.♗xd6 ♔xd6
seizes the opportunity to attack his kingside 64.♔f6 , etc. Instead, Black sets his hopes
pawns. on counterattacking against a3. ]
♗f7 62.♗c5+! QUESTION: Why is this good?
[ 57...gxh5 58.gxh5 leaves the h7-pawn ANSWER: It worsens the position of the
very hard to protect, and Black would be black king.
f o r c e d i n t o t h e h o r r i b l y p a s s i v e ♗g8 QUESTION: But how?
. Rather than that, Black prefers to defend ANSWER: Well, where are you going to go?
a pawn on g6, but either way, his position [ QUESTION: What about meeting
is becoming progressively more 62.♗c5+ with ♔e8 ?
uncomfortable. ] A N S W E R : T h e n I h a v e 63.♔f6
58.♔g5 Now he has to take on h5 anyway. , and Kg7 will decide. ]
gxh5 59.gxh5 ♗g8 60.h6 Fixing the pawn [ QUESTION: Well, OK then, I will try
on h7 and thus ensuring that the black 62.♗c5+ ♔f7 .
bishop will be tied down on g8 for the ANSWER: That has the same problem as
foreseeable future. the move in the game – it blocks the action
QUESTION: But I still don't see how White of Black's bishop, and allows 63.a4 . ]
is going to win. 62...♔e6 63.a4! Exploiting the fact that the
ANSWER: Nor do I, and, frankly, I don't black king has blocked its bishop from
suppose Bronstein did yet, either. But he is defending the knight.
making steady progress and has seized some a6 64.a5! Now the pawn on a6 is another
further plusses, in the shape of creating a weakness. White threatens Be4 and Bb7,
new weakness on h7 and tying down the hence Black's next move.
enemy bishop. Now he will consider how to ♔e5 65.♗f8! QUESTION: Now what?
progress further. ANSWER: White prepares a check on g7.
b5 Black tires of waiting passively and tries Then the black king will not be able to cover
to set up some counterplay with ...Nc4. both f6 (to stop Kf6-g7) and e4 (to stop Be4-
QUESTION: What happens if he waits? b7). The end is near.
ANSWER: Well, as I say, there are various ♘e3 66.♗g7+ ♔d6
possibilities, but the one that probably [ T h i s a l l o w s B e 4 , b u t a f t e r 66...♔d5
provoked the text is the idea of a4, followed 67.♗c3 White threatens Kf6-g7. ]
by Bd4 and a5. That would create a parallel 67.♗e4 ♘d5 The only way to keep the
situation on the queenside to what we just enemy bishop from b7. Black is clinging on
saw on the kingside, but this time, Black desperately by his fingertips, but White will
would not have the pieces with which to inevitably find a way to stamp on his fingers.
defend the queenside. 68.♗f8+ ♔c7
61.♗d4 ♘c4 QUESTION: Doesn't that lose [ Q U E S T I O N : W h a t a b o u t 68...♔e6
a7? , keeping the white king out?
[ ANSWER: The trouble is, Black cannot ANSWER: Then simply 69.♗f5+ ♔e5
afford the time for 61...a6 because of 70.♗c8 ♘c7 71.♗g7+ ♔e4 72.♗a1
62.♗c5 and there is no defence to the threat and again, Kf6-g7 will follow. Black is
of taking on d6 and entering on f6-g7 with unable to defend the two far-flung
which would exchange off the knight on f6, his temporary development advantage.
which defends d5. 11.♗c4 ♘c6 12.h3 ♗e6 13.♗xe6 fxe6
QUESTION: But the other black knight can QUESTION: So what is going on here? As
just recapture. well as a backward d6-pawn, Black has
ANSWER: Yes, but in general, in these ...e5 doubled e-pawns. Surely this must be bad for
structures, Black is relying on dynamism, him?
and so he prefers to avoid too many ANSWER: Actually, not! Once again, it is
exchanges. The latter would tend to suck the our old friend, the 'unexploitable weakness'.
life out of the black position, and could The pawns on e6 and d6 look as though they
eventually see his static weaknesses assume should be weak, but White is not able to
greater importance. The two black knights attack them successfully. How can he attack
can both control d5, whereas the white e6, for example? Only by Qc4, but that is
queen's bishop cannot, so it makes sense to easily met by ...Qd7, and then the white
prevent the exchange of bishop for knight. queen will soon be driven from its exposed
QUESTION: OK, but even so, 7...h6 does position on c4. Likewise, the d6-pawn can
not develop a piece. Can't White exploit this? be attacked by Rad1 and Nb5, but Black can
ANSWER: Not really, because the position defend it by ...Ne8, if necessary, and then ...
is actually not that open. White has no a6 will drive off the enemy knight, after
immediate pawn levers and cannot prevent which White is not able to get any more
Black completing his kingside development pressure against d6.
in peace. QUESTION: It is all very confusing! How
8.♗e3 ♗e7 9.a4 0-0 10.0-0 ♘b8! are we supposed to know when a 'weak' pawn
QUESTION: What on earth is he doing? is weak and when it isn't?
Why undevelop a developed piece? ANSWER: Well, I never said chess was an
ANSWER: Bronstein was very pleased with easy game, but the issue is one of specifics.
this move and it became quite trendy in such A formal weakness, such as a backward or
positions. His point is that the knight on d7 doubled pawn, is only a weakness if it can be
is not ideally placed, as it blocks the queen's effectively attacked. As it is, the black
bishop, which would like to come to e6, central pawn formation here also performs a
guarding the vital d5-square. In addition, the very useful function. If we look at the four
queen's knight itself would be more active on main central squares, e4, d4, d5 and e5, we
c6. see that Black controls or occupies three of
QUESTION: But what about Black's them, whilst his knight on f6 attacks the
development? He is already a long way other one (e4) and forces White to defend it.
behind, and this makes things even worse. Furthermore, Black is constantly threatening
ANSWER: Yes, but once again, the issue is to advance ...d5, given the chance, and
levers. An inferiority in development can be White must always stop this.
serious, even fatal, if the opponent can Annotating the game in the October 1975
exploit his superior mobilization to launch a issue of 'CHESS ', Bronstein makes the
quick attack. But here, we have quite a static interesting observation that "Sax felt that
structure. White does not have any way to Black's central pawn position might prove
lever open lines, so he is not able to exploit too rigid. Curiously, the effect is to make
[ QUESTION: Why didn't he protect the g4- Black will just double rooks on the c-file and
pawn with 25.f3 ? White is unable to avoid loss of a pawn,
ANSWER: As Bronstein points out, that since g4 is now hanging.
w o u l d b e a mi s t a k e , b e c a u s e o f ♘xe3 30.f3 ♖dc8 31.♘a2 d5!? QUESTION:
26.♖xe3 d5! with the threat of 27...Bc5. ] Why didn't he take the pawn?
25...a6 [ ANSWER: Bronstein gave his last move
[ Q U E S T I O N : B u t i s n ' t 25...♕xg4 an exclamation mark, claiming that he
just winning a pawn here? rejected 31...♖xc2 32.♕b1 "because I feel
ANSWER: No, White replies 26.♗xh6! . I can get more out of the position".
( S t r a n g e l y , B r o n s t e i n g i v e s 26.♖g3 Actually, his instinct was correct, but he
instead, but then after ♕h5 Black keeps his should have taken the pawn the other way,
extra pawn. )] as recommended by the computer. ]
26.b3 [ I assume he rejected this because of
[ QUESTION: He seems determined not to 31...♘xc2 32.♗xb6 , but now d5!
defend his centre with 26.f3 . is tremendously strong, 33.-- ( the nasty
ANSWER: Yes, Sax was obviously point being that 33.exd5? opens the black
reluctant to weaken the dark squares in this queen's line of defence of the knight at c2,
way, until absolutely necessary. But he thus losing a whole piece after ♖xb6 .;
should probably have done so here. 33.a5 is relatively best, but after d4
Bronstein then gives ♕f7 27.b3 ♘a5 White's position is hopeless, as his queen
28.♗e3 ♕c7 , which is still about equal. ] is boxed in on a2 and his bishop is shut out
26...♘a5 27.♗e3 ♘c6! This is the difference. of play on b6. Black will infiltrate on the
Black is able to exploit the dynamics, to kingside dark squares with ...Qg5,
obtain a more active position than in the combined with a timely ...d3, ...Rxb6 and
previous note. ...Bc5+. )]
28.♕b2 32.♘xb4
[ The point is 28.♗xb6 ♘b4 29.♕b1 [ White's last chance, not mentioned by
♘xd3 30.♗xd8 ♘xf2 (Bronstein), Bronstein, was 32.c4! dxc4 ( 32...dxe4?
( although the computer suggests that 33.♘xb4 ♗xb4 34.♖d8+ ♔h7 35.♕xe5
30...♘f4 is even stronger, since White is l e t s W h i t e r i g h t b a c k i n t o t h e g a m e)
defenceless after 31.♗a5 ♕xg4 32.g3 33.bxc4 ♖xc4 ( or 33...♘xa2 34.♕xa2
♖xc3! 33.♗xc3 ♘e2+ 34.♔g2 ♘xc3 ♖xc4 35.♗xb6 ) 34.♘xb4 ♖xb4 35.♕xe5
35.f3 ♕g5 36.♕c1 ♘xd1 , winning. )] when White is still on the board. Instead,
28...♘b4 29.♖3d2 QUESTION: Things after the text, he goes down without
seems to have turned against White suddenly. much further resistance. ]
[ ANSWER: Yes, his position has become 32...♗xb4 33.♖e2 d4
extremely unpleasant. The computer [ Perhaps predictably, the computer prefers
actually thinks he should jettison material to grab a pawn by 33...♗c3 34.♕a3 dxe4
with 29.♗xb6 ♘xd3 30.♖xd3 , when Black should win easily enough,
, although Black remains clearly better after but Bronstein commented that "I prefer to
♖d7 . ] eliminate with one pawn move all four
29...♖c6 Now White is in dreadful trouble. hostile pieces". ]
to Bronstein's board, and saw the very to Grandmaster Matthew Sadler for his
position he had been shown in the latter's insights into this game, as he plays the
hotel room that morning. Unfortunately, in Najdorf himself and is an expert on this
the video, he does not give the exact position, whole variation. It seems that this is where
but further on, I have pointed out which things start to go wrong for Black.
position I suspect it was. [ These variations are bewilderingly
e6 7.f4 ♗e7 complex, and theory took a long time to
[ 7...♕b6 is the so-called Poisoned Pawn, establish the best lines for both sides, but
which is the most popular variation at the nowadays, 15...♕e7 is regarded as slightly
time of writing, but Browne's choice was more accurate.
always the main line, and was his own QUESTION: What is the difference?
regular preference. ] ANSWER: Well, the detailed variations are
8.♕f3 ♕c7 9.0-0-0 ♘bd7 10.g4 far too long and complex to delve into here,
[ 10.♗d3 is the alternative, played in the but the main general point is that on e7,
Spassky-Fischer match of 1972, but then the queen defends e6 and f7, whereas from
Browne himself was largely responsible d8 it does not. That makes a difference in
for developing the reply h6! which is various lines, but none but Najdorf buffs
regarded as best for Black, and gives him need worry their heads over exactly what
a decent position, according to theory. As difference. ]
a result, attention shifted back to th e 16.♖g1
older move in the game. ] [ M a t t h e w t e l l s m e t h a t 16.h4!
10...b5 11.♗xf6 ♘xf6 QUESTION: Is this is now regarded as stronger, since the h-
forced? pawn can become quite a dangerous
ANSWER: Almost. attacker in various lines. ]
[ 11...♗xf6?! is known to be good for 16...h6?! Matthew's comment was that this
W h i t e a f t e r t h e s a c r i f i c e 12.♗xb5 was the biggest error of all. Black's last
(the point being that d6 is undefended), ] move is simply too weakening.
[ whilst Fischer's old hobby-horse 11...gxf6 QUESTION: So what should Black do about
has never actually been refuted, to the best his bishop?
of my knowledge, but is not nowadays [ ANSWER: He should retreat at once with
regarded as terribly good. ] 16...♗f6 . Then after 17.fxe6 both 0-0
12.g5 ♘d7 13.f5! The most aggressive. ( and 17...g6 have been extensively tested,
QUESTION: But losing a pawn with check? with results that are not too bad for Black. )]
ANSWER: Yes. White gambits a pawn, to 17.fxe6 g6 18.exf7+ ♔xf7 19.♕e2 ♔g7
open the g-file and lure the enemy bishop to 20.h4!
an exposed square. [ Up to this point, Black had been smashing
♗xg5+ out his moves at blitz speed, as Browne
[ Again, nowadays accepted as best, was still in his book. The previous year, he
although a great deal of research was spent had reached this position, against the
establishing that the alternatives 13...♘e5 ] young American GM, Patrick Wolff. That
[ and 13...♘c5 were less good. ] game had continued 20.♘d5 ♖f8 21.♕g2
14.♔b1 ♘e5 15.♕h5 ♕d8?! I am indebted ♖a7 22.♕g3 ♔h7 and Black was OK and
both with 23...♗f6 runs into 24.♖xf6! [ Retreating the bishop is not a lot better
♕xf6 25.♘d5 ( 25.♕xh6+? ♔g8 than in the previous note: for example,
is much less clear) 25...♕f8 and now there 27...♗d7 28.♖b7 ♖d8 29.♗h3 ♗xh3
is a beautiful finish. 30.♖xe7+ ♔h8 31.♖xg6 and it is all
EXERCISE: Can you find a smart mate over. ]
here? [ QUESTION: But what about 27...♗c5
ANSWER: 26.♕xh6+! ♕xh6 , hitting both white rooks?
( actually, Black can delay the mate a few ANSWER: That is much the best try, and
moves with 26...♔g8 , but it is still mate in the only way to stay on the board. After
five: 27.♖xg6+ ♔f7 28.♖g7+ ♔e8 28.♖b7+ ♔g8 29.♖h1 ♗e6 30.♖xh6 ♗f7
29.♘c7+ ♔d8 30.♕g5+ and mate next Black is a pawn down, but he is holding
move ) 27.♘f6# is a beautiful mate. ] the remains of his position together and
24.♕xe5 ♕e7 his bishops offer some defensive chances in
[ The exchange of queens brings no relief the ending. ]
to Black's position, but the alternative 28.exf5!
24...♖e8 is no better after 25.♕d5 ♗c8 [ M u c h s t r o n g e r t h a n 28.bxc3 ♗c5
26.♗d3 , with Rf1 to follow. ] with excellent drawing chances. ]
25.♕xe7+ ♗xe7 QUESTION: Surely the 28...♖e3 29.♗d3 QUESTION: Opposite-
queens coming off must help Black, since coloured bishops in the ending are supposed
his exposed king should now be less of an to be a drawing factor, aren't they?
issue? ANSWER: Pure opposite-coloured bishop
ANSWER: One would think so, but his king endings have a high drawing factor, yes, but
still proves a millstone round his neck, as the when there are other pieces on the board,
white rooks are so much more active and this effect is much reduced. Here, we still
Black's bishops lack stable outposts. This is have something more akin to a middlegame,
an instructive example of the white initiative with the black king still subject to attack.
extending into the endgame. We have seen The opposite-coloured bishops actually
several examples of this phenomenon already strengthen that attack, by enabling White to
in this book. bring an extra piece to bear on the light
26.♖c6 ♖hc8! squares.
[ This tactical resource is the only way to ♗c5
avoid immediate loss of material, since [ 29...g5 looks as though it drops a piece
after 26...♗d7 the bishops run into lethal after 30.f6+ , but, paradoxically, this was
trouble: 27.♖c7 ♖hd8 28.♘d5 ♗g5 ( or Black's best chance. After ♖xd3 31.cxd3
28...♗d6 29.♖xd7+ ♖xd7 30.♘f6+ ) ( 31.fxe7 ♖e3 is nothing for White)
29.♗h3 and Black is losing a piece, 31...♗c5 32.♖b7+ ♔g6 33.♖f1
( but not 29.♘b6?? ♗e3 . )] White's extra exchange should win, but
[ Similarly, 26...♗c8 27.♖c7 ♖e8 28.♘d5 Black's passed kingside pawns offer some
is no better. ] hopes of defending the position. ]
27.♖b6 ♖xc3? QUESTION: What is this? 30.♖bxg6 Now it is much simpler, as the
Desperation? black king is still in serious trouble.
ANSWER: Exactly. ♖ae8 31.a4 bxa4 32.f6! A neat conclusion.
ANSWER: With White's knight on e2, I like ANSWER: I think the popular view of the
the idea of getting a black pawn to e5 as Stonewall is just incorrect. The opening is
soon as possible, so that the pawn not bad at all, but it does require careful
'dominates' the knight, stopping it jumping to handling and deep understanding.
d4 or f4. On the other hand, it seems to me Unfortunately, there are many relatively
that the knight is quite well-placed on e2 weak players, who have used the line over
against the Stonewall, as it can transfer to f4 the years, usually hoping they can simply
and d3, eyeing up the key e5-square, and it give mate by ...Qe8-h5, etc. The trouble with
also leaves the white f-pawn free to advance such a one-dimensional approach is that it
to f3, supporting the break e4. On the other does not really work against decent white
hand, Botvinnik had forgotten more about play, and there have been numerous games
the Dutch than I will ever know, and he had where Black has suffered a fiasco in such
also played 6...d5 in Game 16 of the same positions.
match, so he clearly believed in it. We have all seen many examples of
7.♘ge2 c6 8.b3 ♘e4 nightmare bad Stonewall bishops, etc, and
[ In Game 16, Botvinnik had played 8...♗d6 the opening has gained a bad reputation as a
9.0-0 ♕e7 , a typical regrouping in such result. But when it is handled by a seriously
positions, and had achieved a satisfactory strong GM, who knows what he is doing, the
position. The game was eventually drawn positions are perfectly playable and also
after a gruelling fight, with Black close to highly interesting. For an excellent
a win in the endgame. explanation of some of the fine points of the
QUESTION: So why did he deviate now? opening, I can recommend the section in the
ANSWER: It is not clear. Possibly he 2012 book, 'Study Chess with Matthew
feared an improvement, although it is not Sadler'.
really the type of position where a strong 11.♕d3
novelty is likely – the position is about [ Kasparov, quoting Boleslavsky, gives
patient manoeuvring. The fact that 11.f3 as more accurate, with the idea of
Botv inn ik dev iates f rom a succes sfu l ♘xc3 ( or 11...♘d6 12.♕d3 and again, a
previous game says quite a lot about his quick e4 ) 12.♘xc3 followed by e4. This
character and objectivity, I suspect. Even supports my contention that the Stonewall
when an opening had gone well for him, he may not be the most accurate reaction to
would analyse the game closely and still White's e3 and Nge2 set-up. ]
look for improvements. ] 11...g5!? QUESTION: Wow! He is really
9.0-0 ♘d7 10.♗b2 ♘df6 QUESTION: So going for it, isn't he?
how do you assess the position? ANSWER: Funnily enough, this is not quite
ANSWER: It is a typical Stonewall, where the attacking gesture it looks. Black has a
White h as a s light advantage, p erhaps , more subtle, positional idea in mind – he
thanks to the weakness of e5, but where wants to play ...g4, so as to secure control of
Black is very solid. the e4-square and stop White organizing f3
QUESTION: I thought the Stonewall was and e4.
supposed to be bad, so how come White QUESTION: So is it any good?
doesn't have the advantage? [ ANSWER: It is interesting, but Kasparov
recommends instead a nice way of and create a weakness in the black queenside.
exploiting White's slight inaccuracy on the a6 19.a4 ♕e7 20.♖ab1
previous move: 11...♗d7! . [ The immediate 20.b5 was possible, but
QU ES TI O N: H u h , t h at d o es n o t lo o k Bronstein prefers to make another
terribly impressive or striking? preparatory move, giving Black the chance
ANSWER: Maybe not, but there is a subtle to take radical measures on the queenside. ]
point to it. The idea is to meet White's 20...b5!? QUESTION: That is certainly
intended 12.f3 with ♘xc3 13.♘xc3 ♗e8 radical! But is the cure worse than the
14.e4 fxe4 15.fxe4 ♗g6 . Now the pin on disease?
the e4-pawn is awkward for White, and ANSWER: It is one of the standard ways of
after 16.♕e3 there is the further clever meeting the mino rity attack , bu t is no t
point dxc4 17.bxc4?! ♘g4! and suddenly, always open to Black. Basically, for this to
White's position is a bit hollow in the centre. work, he needs to be able to prevent White
His queen is attacked, he must guard d4, from playing e4 (for a classic example of
and ...Bg5 and/or ...e5 are in the air . what can happen if he does not, see the final
Black would have good counterplay. ] game of the Karpov-Spassky, Candidates
12.cxd5 Semi-final of 1974). Here, Black has a firm
[ White might have done better to dispense grip on e4 (at least, at the moment), so in
with this exchange and to play the principle, conditions ought to be right for ...
immediate 12.f3 , with a continuation b5. However, it is committal, and one of the
similar to the game. ] other key aspects of successful application of
12...exd5 13.f3 ♘xc3 14.♗xc3 g4 s u ch a s t r u c t u r e , n amel y t h e ab i l i t y t o
This is the point of Black's idea, begun at transfer a black knight to c4, is not present
move 11. He keeps fighting to undermine here.
White's control of e4. [ Kasparov indicates that he would have
15.fxg4 ♘xg4 QUESTION: So how do you preferred 20...♗d7 , with the idea of 21.b5
assess this position? axb5 22.axb5 ♘g4 , and the possession of
ANSWER: The computer likes White a lot, the open a-file gives Black reasonable
but I think his advantage is only relatively play. ]
small, an assessment Kasparov confirms. 21.♗g2 ♘g4 Seizing the chance to transfer
White has weaknesses on e3 and e4, Black the knight to a better square.
one on f5 and the draughtier king. 22.♗d2 ♘f6 23.♖b2 QUESTION: What is
16.♗h3 the idea of this?
[ The computer prefers 16.♘f4 , but that ANSWER: White has the plan of seizing the
would allow ♘f6 and a later ...Ne4, which a-file, by playing Ra1, to threaten axb5.
Bronstein instead hurries to prevent. ] QUESTION: And how is the assessment
16...♘h6 Now the knight is much less well- now?
placed on the edge of the board, and tied to ANSWER: Black is holding the balance. His
defending f5. knight is now on a much better square than
17.♘f4 ♗d6 18.b4! QUESTION: What is before, with the further possibility of coming
this, a minority attack? to e4 and thence possibly to d6, always a
ANSWER: Just so. White aims to play b5 good square for a black knight in these
the classic Stonewall weakness on e5, and points out the variation 36.♖f7 ♗xf7
Black's game is now extremely hard to 37.♘xf7+ ♔g7 38.♘xd6 ♕f3 39.♗g3
defend. ♖a8 , which the computer confirms is
♗e8 only a draw.;
[ 31...♗xe5 32.dxe5 looks terribly B2) Instead, Bronstein recommended
dangerous, but now a blow of fearful 36.♗g3 "when there is no defence
strength demolishes the black position. ] against 37 Rf7". Kasparov queries this,
[ EXERCISE: Can you find a good c i t i n g t he l i n e ♖d8 37.♖f7 ♕g5
positional move after 31...♗e8 ? 38.♖f4 ( but here, Fritz finds the
Looking at the position, one can see that improvement 38.♗f4! followed by 39
the white dark-squared bishop on e1 Qe6, which wins at once; so Bronstein's
would really like to get into the game, so 3 6 B g 3 s e e m s t o w i n i n t h i s l i n e)
as t o ex p l o i t B la ck 's n u me r o u s d ar k - 38...♕g8 .;
square weaknesses. So, Bronstein played: ] B3) Nevertheless, Kasparov is right that
ANSWER: 32.g4!! A lovely positional blow. the best response of all to 35...Qf5 is
With White's pieces mainly infiltrating down the typical computer move, 36.♕a2!!
the queenside, Bronstein suddenly opens a when the queen suddenly threatens to
second front on the other wing. Now his penetrate down the a-file in various
bishop on e1 enters the game with decisive lines. For example, ♗f8 37.♘f7+ ♗xf7
effect. 38.♖xf7 ♕d5 39.♕a7 winning.;
fxg4? With time-trouble adding to his other B4) 36.--;
woes, Botvinnik collapses. C) 35...-- ]
[ As Kasparov points out, the computer 35.dxe5 ♗xe5 36.♖f1 Now, on the other
suggests Black is still hanging on after the hand, the game ends prosaically.
co ld - b l o od ed 32...♗xe5 33.dxe5 ♖g7 ♕g8
34.♖xg7 ♕xg7 35.gxf5 ♕xe5 36.♖a7 [ QUESTION: Is 36...♕h6 any better?
♔g8 , ] ANSWER: Unfortunately not. Then
[ w h i l s t t h e i m m e d i a t e 32...♖g7 37.♖f5! leaves the black bishop without a
is also more tenacious than the text. ] safe square on the long diagonal, as ♗g7
33.♗xe4 dxe4 34.♗h4 ♖xe5? 38.♖xg7! ♕xg7 ( 38...♔xg7 39.♕b2+
Another mistake, which soon ends the game. ♔g8 40.♕a2+ ♔g7 41.♕a7+ wins )
[ Once again, the computer shows the 39.♗f6 decides. ]
toughest defence as 34...♖xa7 35.♖xa7 37.♗g3! Fatally luring the bishop off the long
, and now two queen moves are possible: diagonal.
A) 35...♕h6 is one, after which 36.♗g3 ♗g7
♖d8 37.♗f4 ♕f6 38.♖f7! [ 37...♗xg3 38.♕c3+ mates, ]
( much stronger than Kasparov's 38.♕c2 ) [ as does 37...♕xb3 38.♖f8+ ♕g8
38...♗xf7 39.♘xf7+ ♔g7 40.♘xd6 39.♗xe5# . ]
is decisive.; 38.♕xg8+
B) Kasparov gives the other queen move, [ After 38.♕xg8+ ♔xg8 , 39.♗xb8 wins.
35...♕f5 , as best of all. A beautiful game by Bronstein, and one
B1) Bronstein also considers this, and which set up a thrilling finish to the match.
seems to have lost back his extra pawn. the a8-rook, whilst also stopping the thrust
ANSWER: Yes, but he hangs on to the e5-e6. Since 8...Bb7 allows 9 e6, as we have
pawn on c4, which prevents White's seen, he sacrifices a tempo, to play ...Qd5,
king's bishop from developing to an then ...Bb7 and ...Qd7.
aggressive position on d3. This line had a QUESTION: But I don't like sacrificing
brief vogue of popularity at the time, and tempi in the opening!
is still regarded as probably Black's best ANSWER: Maybe not, but you do like
response to 4 Nc3. ] pawns! As the Roman poet Virgil observed,
8.bxc3 ♕d5 QUESTION: Wow! That looks there is no such thing as a free lunch, and
a bit strange. Black has no pieces developed that applies on the chessboard as well. Black
at all, yet he brings out his queen. What has nicked a pawn and hung on to it, so he
happened to basic opening principles? must expect to have to suffer a bit. If it were
ANSWER: Well, such principles are possible to win the pawn without any
important, of course, but specifics always inconvenience at all, then nobody would
have to predominate. White had a direct offer the gambit with 4 Nc3, of course.
threat of 9 axb5, exploiting the pin against 11.♗a3! Another accurate move. Before
the undefended rook on a8, so Black needed Black has the chance to play e6, White puts
to meet this. He has a very limited choice of his bishop on the a3-f8 diagonal, thus
ways to do so. ensuring that ...e6 will involve the loss of
[ The only real alternative was 8...♗b7 castling rights for Black.
, but this allows White a classic blockading e6
sacrifice with 9.e6 , offering another pawn, [ QUESTION: Is 11...g6 possible? It looks
with the aim of gumming up Black's more attractive than losing castling rights.
development. Three years later, Bronstein ANSWER: It is possible, but after 12.0-0
had a game with this very line, which ♗g7 13.♖e1 Black once again has to
continued fxe6 10.♘g5 ♕d5 11.♕g4 reckon with the threat of e6, shattering his
♕f5 12.♕g3 and now Black felt obliged to pawn structure.
return one of his extra pawns to douse the Q U E S T I O N : O K , b u t t h e n I p l a y ♗d5
white initiative. After e5 13.♕xe5 ♕xe5+ stopping it.
14.dxe5 ♘d7? ( 14...♗d5! holds the A N S W E R : Y e s , a n d n o w 14.♘g5!
balance ) 15.♘e6 ♖c8 16.axb5 axb5 is unpleasant for you. Again, e6 is a threat,
17.♖a7 White had a large advantage, and and e6 leaves you unable to castle and with
went on to win in D.Bronstein-I.Chikovani, serious weaknesses on the kingside dark
USSR Team Championship 1967. ] squares. I continue 15.♘e4 and you have
9.g3 With the black queen exposed on d5, it some very unpleasant problems to solve.
makes sense to fianchetto White's bishop, QUESTION: So was there no other
setting up a threat of Bg2 and then Nh4. alternative? ]
♗b7 10.♗g2 ♕d7 QUESTION: This just [ ANSWER: The computer suggests
loses a tempo, though, doesn't it? 11...♗d5 12.0-0 ♘c6 , developing the
ANSWER: Yes, but Black had to meet the queenside, whilst postponing a decision
threat of 11 Nh4. Essentially, 8...Qd5 was a about the other wing. That makes some
temporary measure – Black wanted to defend sense, but I do not think it really solves the
issue – it merely postpones it. After simply QUESTION: So why do you think
13.♖e1 g6 14.♘g5 , I am not convinced Bronstein didn't play it?
that much has changed. ] ANSWER: I am not sure, as he never
12.♗xf8 ♔xf8 13.0-0 QUESTION: So what annotated the game, to the best of my
is your assessment of this position? knowledge. He may have been worried
ANSWER: I rather like White's game. He about ♗xg2 15.♔xg2 ♕d5+ , when the
has excellent compensation for the pawn, in knight cannot come to e4 and will
the shape of more space, better development, apparently be driven back to a lesser
and some potentially serious dark-square square after ...h6.
weaknesses in the black camp, notably c5. However, here White can calmly offer the
QUESTION: So are you saying that the queen exchange with 16.♕f3 ♕xf3+
whole variation is just bad for Black? ( 16...h6? 17.♘xe6+ wins ) 17.♔xf3
ANSWER: I would not go that far. Firstly, when he still has a clear advantage – he
he has the later Korchnoi idea 7...e6 8 axb5 t h r e a t e n s t o t a k e o n b 5 , w h i c h B la c k
Nb6, which I think is better than the game c a n n o t r e a l l y p r e v e n t , e x c e p t b y b4
continuation. And, secondly, even the game , a f t e r w h i c h 18.cxb4 ♘c6 19.♖ac1
position may not be that bad for him, but I ♘xd4+ 20.♔g2! just looks good for
think it takes a certain kind of player to White, ( but not 20.♔e4? ♘b3
defend such positions. Korchnoi was always , which allows Black tricks, based on a
a notorious pawn-grabber, and won many knight fork at d2. )]
games by nicking hot-looking pawns and Bronstein's move avoids this queen exchange
somehow defending the threats, but it is not variation and just prepares to push the f-pawn,
a strategy at which many players excel. And, which will give a fierce attack.
as we will see, even he could come badly 14...c6? QUESTION: A strange move!
unstuck on occasion. So I would not ANSWER: Yes, I am not sure what Korchnoi
recommend this line to Black as a practical was thinking here. After the text, which kills
proposition, even if the computer claims that his queen's bishop and knight at a single
he is objectively OK. stroke, White just sweeps his opponent away
g6!? QUESTION: Goodness me! You were on the kingside.
talking about dark-square weaknesses, and [ I think the only way to play is the
now he play s that! Surely this mu st be c o m p u t e r ' s 14...♗xg2 15.♘xg2 b4!
terrible? , seizing the chance to disrupt the white
ANSWER: It is certainly very ugly, but it is p a w n c h a i n . N o w 16.cxb4 ♘c6
not easy to suggest a good alternative. Black attacks two pawns, and Black seems to be
needs to find a way to connect his rooks, and holding satisfactorily. ]
there is no other obvious means of doing so. 15.f4 ♕e7 16.♕d2 ♔g7 QUESTION:
14.♘h4?! I am a little surprised by this, I Black seems to have established a blockade
must admit. on the light squares.
[ 14.♘g5 is much more natural, planning to ANSWER: Yes, but Nimzowitsch wrote that
bring the knight to e4 and exploit the dark where an incorrect blockade is established,
squares. I think Black's game would be the opponent will break it down precisely at
pretty difficult then. what appears to be its most fortified point.
That is what happens here – f5 looks to give mate in four, starting with
impenetrable, but... 25.♖xg6+! a n d n o w e i t h e r hxg6 ( or
17.f5! exf5 18.♖xf5! It is as simple as that. 25...fxg6 26.♕f6+ ♔h6 27.♘f5+ ♔h5
The knight fork on f5 means that the rook is 28.♕h4# ) 26.♘f5+ ♔h8 27.♕h6# . ]
immune, so White has succeeded in opening Bronstein's move, although less incisive, is
the f-file and can double rooks next move. motivated by the same idea as 23 Rxe6!,
His attack is crushing. namely the desire to eliminate the defending
♗c8 The bishop prepares to transfer to e6, to bishop. The white knight transfers to f4,
defend f7. from where it threatens not only g6, but also
19.♖f6 ♖a7 20.♖af1 ♗e6 21.♕g5 e6 and h5. The computer confirms that Black
Intensifying the pressure on the dark squares has no defence.
around the black king. 23...♔h8 24.♘f4 ♖e7 25.axb5 axb5
♕e8 26.♕h6 ♘d7 This just loses c6, but Black
[ Note that 21...h6? loses at once. has no moves.
EXERCISE: Can you find a win here? [ QUESTION: What about challenging the
ANSWER: White mates with 22.♖xg6+! queen with 26...♕f8 ?
fxg6 ( or 22...♔h7 23.♖g7# ) 23.♕xg6# . ] ANSWER: That loses at once to
22.♗e4 The last white piece is included in the 27.♘xg6+ ( and even 27.♕xf8 ♖xf8
attack. 28.♘xe6 is good enough. )]
♖g8 23.♘g2 [ Q U E S T I O N : S o , w h a t a b o u t 26...♖g7
[ This is good enough, but the computer instead?
shows a more incisive line, which is also A N S W E R : T h e n a g a i n , 27.♘xg6+!
quite instructive: 23.♖xe6! ♕xe6 24.♖f6 is decisive: fxg6 ( or 27...♖xg6 28.♗xg6
♕e8 . ,etc ) 28.♖f8+ . ]
QUESTION: But what has White gained? 27.♗xc6 b4 Black could have resigned with
He has just given up one of his attacking a clear conscience, but probably want to
rooks for the black bishop. extend the game beyond move 30.
ANSWER: Yes, but he has removed the 28.d5 ♗g4 29.e6! ♕f8 30.♕xf8 ♘xf8
bishop on e6, which was a key defender. 31.h3 Very sadistic.
The important point to understand is one [ Of course, many moves win, including the
which is similar to what we discussed in crude captures 31.exf7 ♖g7 32.cxb4 . ]
the context of Bronstein-Ljubojevic, and is 31...g5
again something Kasparov has emphasized [ 31...♗f5 32.exf7 ♖g7 33.g4 is curtains. ]
when demonstrating his games. All White 32.d6
has done here is exchange off one [ Again, most roads lead to Rome, such as
attacking piece for one defensive piece. In 32.♖xf7 . ]
addition, the attacking rook on f6 has been 32...♖xe6 33.♘xe6 ♗xe6 34.cxb4 ♔g7
replaced by its colleague, whereas there is 35.♗e8 c3 36.♗xf7 Certainly one of the
no replacement available for the black most one-sided hammerings Korchnoi can
bishop that was on e6, where it was doing ever have suffered.
a vital job of defending f7 and f5. And, 1-0
indeed, White can now exploit its absence
He can probably force recovery of the pawn, ♖c8 11.♖c1 ♖c7 12.♕e2 ♕a8
but it will take a tempo or two, and will also QUESTION: Wow, that looks pretty
involve exchanging an extra pair of pieces, extravagant!
which will relieve Black's slightly more ANSWER: Actually, it is a perfectly good
cramped position, so Bronstein decides manoeuvre, despite the result of the game.
instead to avoid such issues. Reti showed back in the 1920s that in such
♗e7 6.♘gf3 0-0 7.0-0 b6 Both sides are double-fianchetto set-ups, arising from Reti's
developing logically. Having dealt with his Opening and similar hypermodern systems,
kingside, Black now brings the queen's the queen can often be effective when
bishop to the long diagonal, neutralizing the backing up the bishop down the long
enemy Catalan bishop. diagonal in this way. Black anticipates later
QUESTION: So has Black equalized? exchanges in the centre, which will open the
ANSWER: Yes, I don't think he has any real long light-square diagonal.
problems here. 13.♖fe1 ♖fc8 This is the other point of
8.b3 ♗b7 9.♗b2 c5 10.e3 QUESTION: It Black's set-up. His rooks double on the c-file,
all looks very cagey stuff by both sides! preparing to give White the hanging central
ANSWER: Well, as we will see, White is pawns, and then target the c4-pawn.
not devoid of ambition in this game. But for QUESTION: So, is this good?
t h e m o me n t , h e k e e p s h i s p o w d e r d r y , ANSWER: It is perfectly reasonable, but
simply developing his pieces and with the white knight on d2, the pawn on c4
maintaining the tension. Such structures are will be better defended than would be the
actually very rich, because the quartet of case were the knight on c3. However, one
central pawns allows many different can imagine the possibility of ...cxd4 and ...
structures to arise. White could take on d5 dxc4, followed by ...Bb4, trying to remove
and/or c5, and Black can do the same on d4/ the knight and undermine the white c4-pawn.
c4. QUESTION: That all sounds quite dangerous.
QUESTION: I always find such positions What is White going to do?
very confusing! ANSWER: Well, the notable thing, if you
ANSWER: I think that is true of many club look at this position, is that Black's heavy
players. The numerous possible pawn pieces have deserted his king – the queen
exchanges mean that one must constantly and both rooks are over on the queenside,
evaluate the consequences of the different whilst White has various pieces, notably his
choices, and many weaker players tend to queen's bishop, eyeing up the enemy kingside.
prefer to clarify the situation, by making That is the clue to what may happen. Black
early captures themselves. But amongst can try to target the white pawns on the
grandmasters and masters, one will queenside, but White will attempt to exploit
frequ ently see the ten s i o n m a i n t a i n e d . the lack of defenders on the kingside. After a
Bronstein understands that taking on d5 slow build-up, the stage is set for a possible
would bring him nothing, and instead, he is explosion.
prepared to recapture with pawns on both d4 14.cxd5 The first central pawn exchange.
and c4, accepting the so-called 'hanging Now Black must decide how to recapture.
pawns'. We will see this later in the game. ♗xd5
♔g7 24.♗xd5 and Black has to surrender [ ANSWER: Believe me, he is. The problem
the queen to avoid a quick mate; is that now White threatens just 22 Rcd1,
B) 20...hxg5 21.♗b2 , with a clear followed by Ng5 and/or Rxd7 and Qe5,
advantage. with mate down the long diagonal. For
QUESTION: I don't understand how all example, 21.♕f4 ♗g7 ( or 21...♕b8
this has happened. One minute, we had a 22.♘g5! , which wins at once, because d5
quiet manoeuvring position, where you is hanging,; and if Black just 'passes', with,
were saying Black was fine, and now say, 21...a5 , then 22.♖cd1 ♗c6 23.♘g5
suddenly he is being mated in variation f6 24.♖xd7! ♖xd7 25.♕xf6 and it is all
after variation. What is going on? over ) 22.♗xg7 ♔xg7 23.♖e7 and 24 Ne5
ANSWER: The thing is that, as I pointed is a deadly threat. ]
out earlier, Black's manoeuvring on the 21...f6 A horrible weakening, but, as the
queenside did leave his king's wing previous note has shown, it is forced. He
somewhat bereft of defenders. Once simply must stop moves such as Ne5 and
things open up, the white pieces suddenly Ng5.
jump out and it only needs one error for 22.♖cd1 Beautifully calm, bringing the last
Black's king to come under a fierce attack. white piece into the attack.
That is what can happen in such c4 The computer confirms that there is
positions. ] nothing better. This move introduces the
19.♕xe4 ♗xd5 20.♕f5!? threat of 23...c3, closing one of White's
[ This works well enough, but the computer deadly bishops, but now the move Black
gives 20.♕d3! as even stronger, when thought his 21st move had prevented turns
White has more than enough compensation out to be playable after all.
for th e pawn, and Black is in serious 23.♘g5! ♖c5 Keene's move defends the
trouble. 21 Ng5 is the threat, and after, for bishop on d5, but it is not enough to hold.
example, g6? ( 20...h6? does not stop the QUESTION: What is going on?
threats: 21.♘g5! hxg5 22.♗xd5 ♕b8 [ ANSWER: The main point is that
23.♕g6! and Black has no way to defend f7, 23...fxg5 now loses to 24.♕d4 , neatly
as his knight does not have a single safe exploiting the drawback of Black's 22nd
square to move to) , there follows simply move, which gave up control of d4: ♗xg2
21.♖cd1 and Black is actually completely 25.♕h8+ ♔f7 26.♕xh7+ ♗g7 27.♕xg7#
lost, as ♗c6 22.♕c3 f6 23.♖xd7 ♖xd7 .]
24.♕xf6 ♗g7 25.♕e6+ ♔h8 26.♘e5! [ QUESTION: But what about taking on g2
is murderous. Once again, we see the instead?
white pieces leaping out of their apparent ANSWER: After 23...♗xg2 there follows
torpor and suddenly creating lethal 24.♖xd7! (eliminating the defender of f6)
threats. ] ♖xd7 ( the computer shows as relatively
20...g6 Once again, the threat of 21 Ng5 best 24...♗h3! 25.♘xh3 ♖xd7 26.♕xf6
leaves Black little choice. c3 27.♗xc3 ♖xc3 28.♕xc3 and White has
21.♕f4 QUESTION: It does not look as an extra pawn, which should suffice to win,
though Black should be in so much trouble although some technical problems remain)
here. 25.♕xf6 when a lovely variation, pointed
The weakness on c4 will soon drop off. ] exd4 17.e5 dxe5 18.♗xf6 gxf6 19.♕g4+
12.f5! Immediately putting his finger on ♔h8 20.♕f5 . ]
Black's error. Now there is a very unpleasant With the text, Najdorf was hoping to create
threat of the pawn pushing on to f6 in some counterplay on the open g-file, but he was
variations, as well as a possible exchange on always an incorrigible optimist, and no such
e6. Black is in trouble. play ever materializes. Instead, he is simply
QUESTION: It seems odd that things should battered on the kingside.
have turned against him so suddenly, with 14.d5 Seizing the chance to relieve the
only a minor inaccuracy on his part. pressure against d4 and drive the knight
ANSWER: To a considerable extent, that is offside (on the previous move, 13 d5 would
the nature of the Sنmisch. In such a sharp have allowed the knight to swing over to the
and doubled-edged variation, the cost of an kingside via e7).
error is always much greater than in a quiet ♘a5 15.♘g3 QUESTION: That is a
position. White has some real trumps here, surprise! Why not take on g7?
in the shape of his bishop-pair and powerful [ ANSWER: That would be equally good.
pawn centre. If he is allowed to exploit these, Bronstein suggests that he did not want to
then Black's position can become critical encourage the black knight to g7, in a line
very quickly. On the other hand, as several such as 15.fxg7+ ♘xg7 16.♘g3 f6
o f t h e g a mes i n m y N i m z o w i t s c h b o o k , but it still seems that White has a decisive
showed, the reverse is also true – White has positional advantage after 17.♗h6 ♖g8
some potentially quite serious defects in his 18.♕e2 . ]
position too, especially his doubled pawns, 15...gxf6
and if he plays too slowly and allows his [ QUESTION: This looks like suicide!
opponent to block up the position, then Surely he had to keep things closed with
things can turn bad for him very quickly, as 15...g6 , didn't he?
well. ANSWER: Yes, that was the only hope,
e5 although objectively, after 16.♗h6 ♖g8
[ 12...f6 13.fxe6 is impossible, so Black 17.♕e2 , White's positional advantage is
tries to close the game. Now he hopes to overwhelming. ]
be able to play 13...f6, but Bronstein does 16.♘f5 ♗c8
not give him time. ] [ 16...♖g8 is slightly more tenacious, but
13.f6! ♔h8 QUESTION: Can't the pawn be t h e n s i m p l y 17.♘e3 kills Black's
taken? counterplay against c4, and White can
[ ANSWER: Not really. 13...gxf6? 14.♗h6 quietly build up with Qe2, Rf3, Bd2, Raf1,
is obviously deadly, ] etc. ]
[ whilst 13...♘xf6 14.♗g5 is also very 17.♕h5 Intending simply Rf3-h3.
unpleasant for Black, in view of the threat ♗xf5 18.exf5 ♖g8 19.♖f3 ♖g7
Ne2-g3-h5. Now taking on d4 leads to a [ Bronstein points out a nice finish after
quick mate after cxd4 15.cxd4 ♘xd4? 19...♘g7 .
( 15...♕c8 is more tenacious, but after EXERCISE: Can you find the win?
16.♗xf6 gxf6 17.♘g3 the position is ANSWER: 20.♕xh7+ ♔xh7 21.♖h3+
obviously dreadful for Black) 16.♘xd4 ♘h5 22.♖xh5+ ♔g7 23.♗h6+ ♔h7
material on the queenside, Bronstein calmly [ Following the improvement outlined in the
returns to his original idea of picking off the next note, the line 6...♗xc3+ 7.♘xc3 d5
remains on g6. became more popular here, when Black
♔h7 30.♘h4 ♕e8 31.♗d4 The knight on cedes the bishop-pair, but can hope to
d3 is again trapped, and Black finally called develop pressure against White's c- and d-
it a day – doubtless a blessed relief, after pawns. ]
what must have been one of the worst 7.♘g3?! QUESTION: This looks natural.
afternoons of Szabo's life. Many readers of Why is it dubious?
this book may know little or nothing of ANSWER: It fails to offer White any
Laszlo Szabo, but if I tell you that he played initiative, as the further course of the game
in both the 1950 and 1953 Candidates' shows.
tournaments, I hope you will get the [ Three years later, in his world
message: he was one of the best players in championship match against Smyslov,
the world in the decade or so after the end of Botvinnik unveiled the improvement:
the Second World War. He simply cannot 7.♘f4 d5 8.cxd5 ♗xf1 9.♔xf1 exd5
have been crushed in such style very often in (all of which had been played in a game R.
his life. Fine-G.Shainswit, New York 1951), and
1-0 now the flank blow 10.g4! , which sufficed
to put the entire line starting with 6...Be7
under a cloud for decades. ]
32 E45 7...d5 8.cxd5 ♗xf1 9.♘xf1 exd5 10.♘g3
Botvinnik,M ♕d7! QUESTION: So what is happening
Bronstein,D here?
17th matchgame, Moscow WCh. (Game 23) ANSWER: Black has emerged from the
[Steve Giddins] opening with a good game. He has solved all
his development problems, and the exchange
1.d4 ♘f6 2.c4 e6 3.♘c3 ♗b4 4.e3 b6 of light-squared bishops has left White with
5.♘ge2 ♗a6 This move, nowadays quite some weaknesses on the light squares in his
standard, was actually a novelty in this game. camp, as well as a rather poor bishop.
Bron stein claims he thought it up over QUESTION: So why is Black's last move so
breakfast on the day of the game – a striking good?
contrast to modern world championship ANSWER: Bronstein was very proud of his
matches, where such novelties are liable to play in this game, and gave a long note to
have been subjected to deep analysis months the last move, explaining how it was "the
in advance. start of a series of strong moves, which have
[ In the 13th matchgame, he had played the aim of confusing the opponent and of
5...♗b7 , but failed to equalize. ] disorganizing his harmonious train of
6.a3 thought".
[ 6.♘g3 is also a popular alternative, and is QUESTION: What exactly does that mean?
nowadays regarded as the best try for an ANSWER: I am not sure. The immediate
advantage. ] point of 10...Qd7 is to stop the white knight
6...♗e7 coming to f5. In the more medium term,
21.f3 QUESTION: Is this good? It weakens completed and his pressure down the e-file
e3, after all. reaches its apotheosis. It is time for the
ANSWER: Indeed it does, but therein lies breakthrough. I am astounded to see that
the dilemma. The advance e4 is White's only Fritz continues to assess this position as dead
real hope of serious counterplay, so he feels equal.
obliged to attempt it, but there is always the f4 30.♘e5?!
risk that, if he fails to get in e4 in [ QUESTION: But the black queen is
favourable circumstances, then f3 will just attacked after 30.exf4 ?
prove to be a weakening, as in fact happens. ANSWER: Yup! Bronstein had a queen
g5 QUESTION: Starting the attack? s a c r i f i c e i n m i n d . H e i n t e n d e d ♕xe1
ANSWER: Not exactly. Bronstein says it is 31.♘xe1 ♖xe1 32.♕d3 ( 32.♕d2 ♘xf4
too early to speak of a direct attack, and that is no better) 32...♘xf4 33.♕a6 ♘f5
Black still needs to bring up his forces. But with a winning position.
the move starts the process, by setting up an QUESTION: So what should White have
eventual pawn offensive on the kingside, as done? ]
well as just defending the h4-pawn, which [ ANSWER: Bronstein himself gives
White's last move attacked. 30.♘f2 as "logical", but assesses the
22.♘d3 ♕e6 Black consistently stops White p o s i t i o n i n B l a c k ' s f a v o u r a f t e r fxe3
playing e4. 31.♘g4 ♗g7 32.♖xe3 ♕f7 , because of
23.a4?! QUESTION: What does that do? the weak d4-pawn and the active black
ANSWER: I assume Botvinnik was disturbed knights, which threaten to jump into f5 and
by the threat of ...a4, followed by ..Nc4, but f4.
this is not a very nice move to play. His QUESTION: Do you agree?
position is passive and unpleasant though, ANSWER: Yes, although the computer
and it is always hard to suggest constructive claims White is just equal here too. I
moves in such positions. don't think this is so, but it is clear that
[ Maybe 23.♖d1 makes more sense, this is in any case what White should
defending d4 and hoping to arrange Nf2 have tried. Instead, Botvinnik jettisons a
and e4, but Black will not sit idly by in pawn in desperation. It was noteworthy in
the meantime, of course. ] this match that whenever the position
23...♘g6 24.h3 f5 25.♗c3 Bronstein writes became sharp and concrete, Botvinnik
that Botvinnik clearly feared ...Rf7, ...Bf6 tended to start going astray, partly because
and ...f4, so he overprotects d4. However, he was usually in time-trouble by then. ]
the problem with this move is that he 30...♗xe5! QUESTION: That is not the way I
relieves Black of any worries over his c7- would have expected Black to recapture! I
pawn, which permits the black queen's rook would have taken with the knight.
to come to the e-file. [ ANSWER: In which case, you would have
[ S v e s h n i k o v ' s s u g g e s t i o n o f 25.♖ac1 fallen into Botvinnik's trap. After
would at least make that plan more difficult 30...♘xe5? 31.dxe5 ♗xe5 32.exf4
to carry out. ] White is winning, because of the pin down
25...♗f6 26.♖e1 ♖ae8 27.♕d1 ♖f7 28.b3 the e-file: for example, ♗d4+ 33.♕xd4
♖fe7 29.♗b2 Black's build-up has been ♕xe1 34.♕h8+ ♔f7 35.♕f6+ ♔g8 36.f5
idea was 17.♕xh6 ♖e8? 18.♖e5? is the only way to avoid immediate
( instead, Kasparov points out that disaster, but just 22.♖xe8+ ♕xe8
18.♗xd3 cxd3 19.♘f5 is winning for 23.♕xh5 is obviously much better for
White, as the computer confirms: ♗f8 White ) 22.♕g7+ ♔e8 23.♕h8+ ♗f8
20.♕g5+ ♔h7 21.♘bd6 ♖xe1 22.♖xe1 24.♘bd6# is a delightful example of
and the rook comes to the third rank what problemists call a double pin-
w i t h d e c i s i v e e f f e c t) 18...♘xe5 mate.;
( b u t t h e n h e r e a l i z e d t h a t 18...♘g4! C2b) QUESTION: So can Black do
simply wins for Black) 19.♘h5 ♘xh5 better after 18 Nf5 then?
20.♕h7+ ♔f8 21.♕h8# .; A N S W E R : Y e s . 18...♗e4
B) QUESTION: So does that mean is the best defence. Now the next
17.♕xh6 wins for White? critical position arises after 19.♘bd4
ANSWER: Actually, no. Black has the ♖e8 20.♘xh6+ ♔f8 21.♕g5
stronger defence ♘e8! , as pointed out . Kasparov devotes half a page of
by the computer and Kasparov (the latter solid analysis to this position, and
presumably with the help of the former). then adds that he does not pretend to
The analysis then continues 18.♖e5 f5 have exhausted the possibilities. But
19.♕g6+ ♘g7 20.♘xf5 ♖xf5 21.♖xf5 his analysis suggests that the best
♕e8 22.♕g4 d5 with an unclear move for Black is ♗c5! , after which
position.; he was unable to find a win. There is
C) However, Bronstein had actually little sense in reproducing all of his
decided over the board on a stronger analysis, but here is one typical line:
continuation of the attack, starting with 22.♖xe4 ♖xe4 23.♘df5 ♔e8
the exchange on d3: 17.♗xd3! cxd3 . 24.♕g7 ♕c7 25.♕xf6 d5 26.♕h8+
C1) N o w h e i n t e n d e d 18.♕xh6?! ♗f8 27.♘g8 ♔d7 28.♘f6+ ♔e6
, b u t t h i s i s a g a i n m e t b y ♘e8! 29.♘xe4 dxe4 30.♘g7+ ♔d7
with unclear play: 19.♖e5 f5 20.♘xf5 31.♕h3+ ♔d8 32.♕f5 ♖c8 33.♘e6+
♖xf5 21.♖xf5 ♘g7 , etc.; fxe6 34.♕xf8+ ♔d7 35.♕g7+ ♔d8
C2) Stronger is 18.♘f5 as given by 36.♕g8+ with a draw by perpetual
Kasparov. check.
QUESTION: Delaying the capture on QUESTION: So, the upshot seems to
h6? be that Bronstein's sacrifice was only
ANSWER: Indeed. To paraphrase one good for a draw?
of Tony Blair's most notorious lines, ANSWER: That seems to be the
"This is no time for platitudes, but I feel conclusion, if all the analysis is
the threat is stronger than the execution. correct. But the position is so
" Now there is a maze of complicated complicated that even with a modern
lines, including some lovely mates. engine, it is hard to be really
C2a) Thus, 18...♖e8? runs into definitive. From a practical viewpoint,
19.♘xh6+ ♔f8 20.♕g5 the sacrifice gives enormous chances.
(simply intending 21 Nf5 and mate) Keres was one of the best players ever,
♗c5 21.♘f5 and now ♖e6 ( 21...♘h5 and a noted defender in his mature
ANSWER: The point is to deny Black time to Golombek was described as being brave or
play ...c5 and then ...Nc6. lacking caution. Bronstein's point is
[ Now 11.a4 would be met by c5 12.a5 p r e s u m a b l y t h a t a f t e r 14.♘f3
with a nice queenside initiative. Black ( but I think instead, Bronstein would have
therefore feels he has to stop the white a- continued 14.♘c2 , keeping his options
pawn advancing to a5, but in order to do so, open ) , B l a c k n o w p l a y s 14...e5
he must obstruct his own c-pawn. ] and the white knight is on the wrong circuit
11...♘c6 Now the c7-pawn is blocked, Black to be able to exploit the d5-square. ]
will find it much harder to establish a central 14.♗g5 The bishop has found freedom and
blockade, and will also have always to now exerts unpleasant pressure on the black
reckon with the thrust c4-c5, undoubling the position.
white c-pawns and trying to open the ♕e7 Golombek's move is part of a subtle
position for the white bishops. idea, and reveals why he has avoided playing
[ QUESTION: What if he stops the a-pawn ...d6; he wants to keep the d6-square for his
i n a d i f f e r e n t w a y , w i t h 11...a5 queen.
, keeping the chance to follow up with ... [ The computer thinks Black has no
c5 and ...Nc6? p r o b l e m s a f t e r 14...h6 , but I imagine
ANSWER: But then a later ...c5 would Golombek was afraid to weaken his
leave the b6-pawn very weak. Black also kingside in this way.
wants to keep the a5-square free for his QUESTION: But where would the bishop
knight, as he would like at some stage to retreat to?
attack the c4-pawn with ...Na5 and ...Ba6. ] ANSWER: The mov e 15.♗h4 is quite
12.e4 And this is the other downside of possible, since Black dare not try to grab a
putting the knight on c6. White has relieved pawn with g5 16.♗g3 exd4? in view of
the blockade on e4 and got his central pawns 17.e5 ♘h7 18.cxd4 , when the d4-pawn is
in motion. immune because of ♘xd4?? 19.♗xh7+ . ]
QUESTION: It looks as though White is 15.♘c2 ♕d6 QUESTION: What do you
getting what he wants? think of this? It looks a bit strange.
ANSWER: Indeed. Black has been outplayed ANSW ER: Yes, i t do es . Th e id ea i s t o
over the past few moves and now has some pressure the white centre, specifically the d4-
problems. pawn, but tactically, the threats do not work,
fxe4 13.fxe4 e5 Black must do something and the queen is not great on d6.
to restrain the white central pawns, and this 16.♗h4! QUESTION: Why the exclamation
is the logical way to do so. However, it does mark? Doesn't the move just lose the pawn
weaken the square d5, which Bronstein on d4?
immediately exploits, but I still think 13...e5 [ ANSWER: No. The d4-pawn cannot be
is the only correct move. taken, because after 16.♗h4 exd4 17.cxd4
[ Interestingly, Bronstein himself, in the ♘xd4? there follows 18.♗g3 ♕c5 19.♗f2
tournament book, comments that: "A less and Black loses a piece. Meanwhile, 16
brave and more cautious player would have Bh4 sets up possibilities of Bg3, pinning
played 13...d6 here". There cannot have the black e-pawn, which could be very
been many occasions when Harry awkward in some lines. ]
the case. After 19...exd4 20.♘d5 ♕d8 26.exd5 ♘xd5 White just takes on e5:
21.cxd4 ♘xd4 22.♘xf6+ gxf6 23.♕h5 27.dxe5 with a crushing position. White is
the attack is too strong: for example, ♖e5 only too happy for the centre to open like
( 23...♕e7 24.♗xf6 ♖xf6 25.♖xf6 ♕xf6 this, because in the resulting position, his
26.♕xe8+ wins ) 24.♕h6 and f6 drops off bishops will be devastating, especially
next move. The black king is too weak to with the black king so exposed to the
survive without ruinous material losses. ] elements. ]
20.♘d5 ♗xd5 21.cxd5 QUESTION: So [ QUESTION: OK, I see that. So maybe
what is the assessment? Black should instead go after the bishop
ANSWER: Black has avoided an immediate with 25...♘h5 ?
catastrophe, and his position is solid enough, ANSWER: A very sensible suggestion, but
but strategically, he is in serious trouble. Br o n s t e i n h a d f o r es e e n t h i s , a n d h a d
His position is cramped, and all his pieces prepared a bone-crunching reply:
passive, whilst White has tremendously 26.♗xe5+!! . Now we see another point of
active pieces and the initiative in all areas his super-subtle 24th move. The following
of the board. His bishops are all-powerful, variation is forced: dxe5 27.♕xe7 ♖xf1+
his pawn centre rock-solid, and he can play 28.♖xf1 ♖xe7 29.♖f8+ ♔g7 30.♖xd8
on both the kingside and also the queenside, and White has a winning endgame. ]
after a4-a5. In fact, Black is just about Golombek's move is more solid, just trying to
strategically lost. shore up his central structure, but the bishops
c6 22.♕b3 ♔h8 23.♖ae1 h6 24.♕a3!! now embark on a de mo lition op eratio n
QUESTION: What is the point of that? I worthy of Fred Dibnah.
don't think it would have entered my head to 26.dxc6 ♘xc6 27.♗b5 As Bronstein points
play this! out, this is what is known amongst chess
ANSWER: I sympathize with you. Bronstein problemists as a 'half-pin': neither of the
himself described it as "a difficult, but black knights is actually pinned, but if either
useful move". In fact, I think it is the best one moves, the other will be.
move of the game, by far. Black's 21st move ♖xf1+
weakened his d6-pawn, and the white queen [ Thus, 27...♘db8? loses to 28.d5
now takes aim at that point. The true when the knight on c6 is pinned. ]
effectiveness of the move becomes clear as [ A f t e r t h e i m m e d i a t e 27...♘cb8
the game unfolds. , t h e r e p l y 28.♖f5 is deeply unpleasant,
QUESTION: But doesn't this move lose the with huge pressure against the black
d5-pawn? centre. ]
ANSWER: Yes, but as we will see, White is 28.♖xf1 ♘cb8 29.♗c4 Now 30 Rf7 is
happy to trade that for the black centre pawn, threatened.
which will open the position completely for ♖f8 30.♖xf8+ ♕xf8
the white bishops. [ 30...♘xf8? 31.dxe5 wins, once again
g5 25.♗g3 ♘d7 QUESTION: Why didn't because of the pressure exerted along the
he take the pawn on d5? I thought that was a3-f8 diagonal by the white queen. ]
the point of playing 24...g5? 31.dxe5 And here, too, the black centre
[ ANSWER: The trouble is, after 25...cxd5 dissolves, thanks to the white queen's
pressure on the a3-f8 diagonal. The depth King's Indian tabiya. Both sides have
and strength of Bronstein's 24th move is developed their pieces within their own
very clear to see. camp.
♘c5 QUESTION: The white bishop is on g2,
[ 31...dxe5 32.♕xf8+ ♘xf8 33.♗xe5+ rather than e2, as is more often seen these
wins the knight at b8. ] days. Is it worse there?
32.exd6 This position is worth comparing ANSWER: No, not at all. Although
with the one after White's 25th. The black variations such as Nf3 and Be2 and the
centre has done a real "here you see it, here S نm i s c h w i t h f 2 - f 3 a r e p e r h a p s m o r e
you don't" act. The devastating strength of common nowadays, the g3 system has always
the white bishops is obvious. been one of the most highly-respected set-
♘xe4 33.d7! A nice combination to finish. ups for White in this opening. It was for
♘c5 many years a favourite with such KID-killers
[ The point of White's play is the variation as Korchnoi and Petrosian, and to this day,
33...♘xd7 34.♕xf8+ ♘xf8 35.♗e5+! many KID practitioners hate playing the
♔h7 and now the sting in the tale: 36.♗d3 black side of this system.
and the knight at e4 drops off. A QUESTION: What is the logic of putting the
beautifully thematic variation, since the bishop on g2?
whole game has been about the strength of A N S W E R : W e l l , t h e S نm i s c h b e i n g
the white bishop-pair, and this position something of an exception, White rarely gets
illustrates it perfectly. ] to attack the black king in this opening;
34.♗e5+ ♔h7 35.♗d3+! rather, Black is the one who usually becomes
[ 35.♗d3+ ♔g8 36.♕a2+ is curtains. An active on the kingside, whilst White's play is
absolute textbook on the exploitation of generally on the other flank. From that
the bishop-pair in the middlegame, and viewpoint, it makes sense to have the bishop
replete with many of the delightful pointing down the long diagonal, towards
combinational touches that characterize the enemy queenside.
Bronstein's best games. ] 8.b3!?
1-0 [ Nowadays, the main line is 8.e4 c6 9.h3
. White's plan appears logical enough, but
as we shall see, the long diagonal is not
35 E68 necessarily the best place for the white
Zita,D queen's bishop. It is usually better placed
Bronstein,D on e3, supporting the key d4-square. ]
Moscow-Prague match, Moscow (Game 26) 8...♖e8 9.♗b2 c6 10.e4 exd4 11.♘xd4
[Steve Giddins] QUESTION: I always have difficulty
understanding why Black makes the
1.c4 e5 The game starts out as an English exchange ...exd4 in such positions. Hasn't he
Opening, but soon transposes into the King's just surrendered his central outpost and left
Indian Defence. himself a weak backward pawn on d6?
2.♘c3 ♘f6 3.♘f3 d6 4.d4 ♘bd7 5.g3 g6 ANSWER: Well, you are not alone in finding
6.♗g2 ♗g7 7.0-0 0-0 This is a typical this difficult to understand. When Black first
started to play this way, most grandmasters use of his chances, but with accurate play by
were extremely sceptical, for just the reasons Black, it is usually possible to generate
you give. It was games such as this and the enough activity to prev ent White from
next, which opened people's eyes to the settling down to the exploitation of the
viability of such an approach by Black. weakness. Black must remain active, and
QUESTION: So what is the point? make full use of all of his possible 'guerrilla
ANSWER: Well, firstly, the exchange opens tactics', by harassing the e4 pawn,
the long diagonal of Black's KID bishop, and threatening tactics along the a1-h8 and g1-a7
the half-open e-file for his rook on e8. So, he diagonals, using the half-open e-file, etc.
has some pressure against e4 and d4. He can ♕b6 12.♕d2?! QUESTION: What is wrong
augment the former with ...Nc5, which with this? It looks very natural.
exploits another advantage of the exchange ANSWER: Maybe, but actually, the move
on d4, namely the freeing up of the c5- weakens the b3-pawn, of all things.
square as an outpost for the black pieces, and QUESTION: You are joking?
e5 is another possible outpost. ANSWER: No, not at all. As we will see,
QUESTION: OK, but White also gets some Black follows up with ...a5-a4 and takes on
moves, too! What is to stop him just piling b3, and then ...Nc5, when the b3-pawn will
up on the d-file and attacking the weakness be attacked.
on d6? [ K a s p a r o v p o i n t s o u t t h a t 12.♕c2
ANSWER: Ah, well now you are getting to was better, defending both b3 and e4, with
the real secret of such positions. In his the point that the knight on d4 is indirectly
famous book on the 1953 Zurich Candidates, defended, thanks to ♕xd4?? ( the other
Bronstein dealt with this question head-on, tactic 12...♘xe4 also fails after 13.♘a4
as it is crucial to the viability of Black's followed by the capture on e4) 13.♘a4
play. On the face of it, it looks the simplest trapping the queen. ]
thing in the world, just to bring a rook to the 12...♘c5 13.♖fe1 a5 This move is standard
d-file and then move the knight away from in the King's Indian, and usually has the
d4, attacking d6. But, as Bronstein explains, primarily defensive function of preventing
the d4-knight is actually a very important the knight on c5 being driven away by b3-b4.
piece and fulfils a number of key roles on Here, however, that is not a threat anyway,
that square. It shields the long diagonal of and ...a5 is instead a more aggressive move,
the enemy KID bishop on g7, it restrains the intending queenside play with ...a5-a4.
possible enemy pawn breaks ...b7-b5 and ... QUESTION: But White is still doing fine
f7-f5, and it covers the e6-square, preventing here, isn't he? After all, he has more space
the enemy bishop from coming to that square and the biggest weakness on the board is still
and attacking the c4-pawn. Consequently, it the d6-pawn.
is not so easy for the knight to move away, ANSWER: Actually, I think White is
and all the while it stays on d4, the d6-pawn already slightly worse. As we have explained,
is not so easy to get at. he cannot really generate any pressure on d6,
QUESTION: So the weakness of d6 is not whilst Black's threats against e4, d4 and b3
important? give him the initiative. White is still holding
ANSWER: It might be, if Black fails to make the position, but he needs to be careful.
[ 21.♗f3 is relatively best, but then simply improved methods against the Sنmisch, and
♘g4+ 22.♗xg4 ♗xg4 leaves ...Bh6 the normal King's Indian move order became
threatened, and 23.♖ee1 ♘xb3 24.♘xb3 more popular.
♗xc3 25.♕xc3 ♕f2+ wins at once. ] 4.♘f3
21...♘cxe4 22.♕f4 [ QUESTION: Wouldn't White stand better
[ Or 22.♘xe4 ♘xe4 23.♕f4 f5 after 4.dxe5 dxe5 5.♕xd8+ ♔xd8
and the pin on the d4-knight costs yet , with Black having lost castling rights?
more material. ] ANSWER: Not really. Various games have
22...♘g4+ 23.♔h1 f5 Once again, the shown that Black can play ...c6, tuck his
deadly pin along the long black diagonal king away on c7, and organize his pieces
wins further material. quite well in this position. White does not
QUESTION: The KID bishop certainly have any real advantage here. ]
proves its worth in this game! 4...♘bd7 5.g3 Now we head back to the
ANSWER: Indeed. One can easily see why same line as in the previous game.
this game was so influential. Who wouldn't g6 6.♗g2 ♗g7 7.0-0 0-0 8.b3?!
want to play the King's Indian after seeing Once again this mistaken development. As
this game? we noted in the previous game, the bishop is
24.♘xe4 ♖xe4 25.♕xd6 ♖xd4 26.♕b8 better placed on e3 in this line, keeping a
♖d8 Black has two pieces and two extra weather eye on the important d4- and f2-
pawns for the rook, and his minor pieces all squares.
exhibit deadly activity. ♖e8 9.e4 exd4 10.♘xd4 ♘c5 11.♖e1 a5
27.♖a8 ♗e5 28.♕a7 ♕b4 29.♕a2 ♕f8! 12.♗b2 a4
A nice geometrical touch. The queen [ 12...c6 was a perfectly good alternative,
switches to h6, as White continues to be torn not hurrying with the advance of the a-
apart on the dark squares. pawn, but the text is also good. ]
30.♗h3 ♕h6 13.♖c1?!
0-1 [ QUESTION: I am confused – I thought the
main point of playing ...a7-a5 in such
positions was to secure the position of the
36 E68 knight on c5. Why doesn't White drive the
Pachman,L knight away with 13.b4 and at the same
Bronstein,D time keep the a-file closed?
Moscow-Prague match, Moscow (Game 27) ANSWER: The problem is that Black has
[Steve Giddins] the unpleasant reply a3! and after 14.♗c1
♘e6 , White's pieces are hanging rather
1.d4 ♘f6 2.c4 d6 3.♘c3 e5 This Old unpleasantly on the long diagonal. ]
Indian move order was quite a popular way [ For that reason, White hurries to move his
to reach the King's Indian in the early days rook from the long diagonal, but Kasparov
of the system's popularity. The idea was to points out that 13.♖b1! was a superior way
avoid the Sنmisch variation, which was of doing this, keeping the bishop on b2
r e g ar d ed as v e r y d an g er o u s f o r B l ac k . defended. The rook does not really do
Subsequently, though, Black found various anything much on c1, whilst the bishop on
19...h4 20.♖d2? Continuing the idea from no hope for White, as the computer
the previous move, but running into a confirms. ]
decisive tactical blow. 24...♕xg3+ 25.♔h1 ♕xc3 26.♖a3 ♗xh3!
[ Kasparov gives 20.♖c2 as "a worse, but Winning nicely. Kasparov gleefully points
still defensible position", ] out that, just as 20...Rxa1! was the first move
[ w h i l s t F r i t z s l i g h t l y p r e f e r s 20.b4 by the black queen's rook, so this decisive
, although Black is still clearly better after blow is the first move of his queen's bishop.
♘cd7 . ] [ The computer (and Kasparov) adds that
Now Bronstein strikes with a combination 26...♖xe4! a l s o w i n s : 27.♗xe4 ♕xh3+
extremely similar to that seen in the previous 28.♔g1 ♕e3+ 29.♔h2 ♕f4+ 30.♔g1
game. The exchange sacrifice on a1 is again ♗g4 31.♕xb3 ♕xd6 , etc. ]
the key, removing the bishop that is the QUESTION: It is amazing how it all fits
linchpin of the white position, whilst the together tactically.
loose units on d4 and b3 are also a key ANSWER: Well, that is the greatness of a
element in the combination. player such as Bronstein. But the whole
20...♖xa1! 21.♖xa1 ♗xd4 22.♖xd4 ♘xb3 thing is of course based on his superb
23.♖xd6 positional build-up, and the looseness of the
[ K a s p a r o v p o i n t s o u t t h a t 23.♖d2 various white units.
is slightly more tenacious, preventing 27.♖xb3 ♗xg2+ 28.♔xg2 ♕xc4 29.♖d4
Black's next, but even so, after simply ♕e6 30.♖xb7 ♖a8 31.♕e2 h3+
♘xa1 24.♕xa1 hxg3+ 25.fxg3 ♕b4 [ There is no defence after 31...h3+ 32.♔g1
Black has an extra pawn and the white ♕e5 followed by ...Ne6. It is safe to say
pawn structure is wrecked. Black should that these last two games did more than
wrap this up without too much trouble. ] any others to popularize the King's Indian
23...♕xf2! QUESTION: Can't Black just take Defence, which went on to become just
on a1 first? about the single most popular defence to 1
[ ANSWER: No! That would have fallen d4, over the next few decades. ]
into Pachman's trap. There would follow 0-1
23...♘xa1? 24.♘d5! turning the tables
completely. Now White is the one who
stands better after ♕xf2 25.♘f6+ ♔h8 37 E69
26.♘xe8 hxg3+ 27.♔h1 ♘b3 Reshevsky,S
( 27...♗xh3? 28.♕xa1+ is check ) 28.♖f6 Bronstein,D
, etc. ] Zurich Candidates (Game 28) 1953
24.♖a2 [Steve Giddins]
[ Bronstein points out the attractive mate
after 24.♕xb3 hxg3+ ( the prosaic 1.d4 ♘f6 2.c4 g6 3.g3 ♗g7 4.♗g2 0-0
24...♕xg3+ 25.♔h1 ♕xd6 5.♘c3 d6 6.♘f3 ♘bd7 7.0-0 e5 8.e4 ♖e8
is also winning) 25.♔h1 ♗xh3! 26.♖g1 9.h3 QUESTION: The same variation as in
♗xg2+ 27.♖xg2 ♕f1+ 28.♖g1 ♕h3# . ] the previous two games?
[ 24.♖d3 is possibly the best, but after ANSWER: Yes, but here we see how theory
♘xa1 25.♕xa1 ♘d7 (Bronstein), there is had advanced in the intervening seven years.
Rath er than b3 and Bb2, which proved why such positions require extremely
unsuccessful in the earlier games, Reshevsky precise handling by Black. It is not a set-
adopts a plan, which is still the main line to up for those who just like to be able to
this day. The early h3 covers the vital g4- play natural, common-sense moves – Black
square, preventing the black knight needs to show really deep positional and
manoeuvre ...Ng4-e5, and also preparing to tactical insight. ]
place his queen's bishop on e3, from where it 12...c6 13.♗e3 ♘fd7 14.♖ad1 a4
defends the knight on d4 and discourages the QUESTION: So what is happening now?
black queen from occupying b6, as it did in White is fully mobilized, whereas most of
the two previous games. Black's pieces are on the back rank. Surely
exd4 10.♘xd4 ♘c5 11.♖e1 a5 12.♕c2 White is better?
[ The main line of development, although ANSWER: Possibly slightly, but not much.
K a s p a r o v s u g g e s t s t h a t 12.♖b1 Chances are fairly balanced. It is true that
is more useful, intending to meet a4 with Black has much less direct pressure than in
13.b4 axb3 14.axb3 c6 15.b4 ♘e6 the preceding examples, but he still has some
16.♘de2 when White retains some ideas. He can bring his queen to a5 and his
advantage. knight to e5, hoping to pressure the c4-pawn.
QUESTION: Why is he better here, when If White defends it with b2-b3, then an
he was not in the two previous games? exchange of pawns on b3 will open the a-file
ANSWER: Because he is better organized for Black, and a later ...Qb4 may exert some
here. The black knight has been driven to pressure on the pawn. By these means, Black
a less-than-ideal post on e6, where it hopes to prevent White settling down to the
blocks the bishop on c8 and rook on e8. exploitation of d6.
White is now quite well-placed to follow 15.♘de2?! QUESTION: Why don't you like
up with Be3, Qd2 and Rbd1, when he really this move? Isn't it logical to unmask the
will have some pressure against the attack on d6?
backward d6-pawn. ANSWER: The idea is tempting in itself, but,
QUESTION: The differences between this as we will see, it does not prove very
and the previous two games don't seem effective.
that great, yet the assessment is reversed! QUES TI ON: S o wh at sh ou ld Wh it e d o
ANSWER: Well, chess is a subtle game. instead?
In general, these positions with ...exd4 ANSWER: A good question!
ar e v er y d o u b l e- e d g ed . T h e f ir s t tw o [ This was something of a 'hot' position in
games showed an ideal scenario for Black, the Zurich tournament, and was reached
where White was never able to generate several times. In Stahlberg-Boleslavsky,
any pressure against d6, whilst White's W h i t e t r i e d 15.f4 , stopping the black
own weaknesses were under constant knight coming to e5.
pressure. But it only takes a few subtle QUESTION: That seems very logical, in
changes to the position for the roles to be view of what you said about Black's plan
reversed – if Black's 'guerrilla tactics' are to attack c4.
n e u t r al i z e d , t h e n t h e w ea k n e s s o f d 6 ANSWER: Yes, but the move also weakens
really will become significant. This is the white e4-pawn. Play continued ♕a5
16.♗f2 , and now ♘b6 , hassling the c4- 15...♕a5! 16.♗f1 This looks rather
pawn from another direction. After 17.♗f1 retrograde, and seriously weakens f3, but
♗d7 Black followed up with ...Rad8 and Black was threatening ...Ne5 and/or ...Qb4,
stood fine. ] attacking the pawn on c4, which cannot now
[ 15.♖e2 was tried by Averbakh, aiming to be defended by b2-b3. Note how White is
double on the d-file, but after ♕a5 16.♖ed2 already missing his knight on d4, which
♘e5 17.b3 axb3 18.axb3 ♕b4 would have been defending b3.
, Black proved to have adequate play. QUESTION: Huh? Isn't d6 hanging?
This position is a good example of what [ ANSWER: It turns out that after
we discussed in Game 26, regarding the 16.♖xd6?! ♘e5 Black is threatening ...
position of the white knight on d4. Here we Nxc4.
see that White is all ready just to move QUESTION: OK, so we defend it with
that knight away, uncovering a double 17.b3 .
rook attack on d6, but the pressure against ANSWER: Yes, and now Black reveals his
b3 prevents the knight from leaving its other threat: axb3 18.axb3 ♗xh3! .
post. ] QUESTION: But...
[ QUESTION: So, what should White play ANSWER: After 19.♗xh3 ♘f3+ 20.♔f1
then? ♘xe1 21.♔xe1 White has two pieces for
ANSWER: Kasparov suggests the the rook, but now follows ♘xe4 22.♖d3
interesting 15.b4!? , challenging the black ♕a1+ and Black is winning: for example,
k n i g h t . A f t e r axb3 16.axb3 ♕a5 23.♘b1 ♖a2 24.♕d1 ♗f6 (covering d8)
White continues to try to force the advance and Black's initiative is too strong.
of his b-pawn: 17.♖b1 . He claims some QUESTION: Lightning from a clear sky!
i n i t i a t i v e f o r W h i t e a f t e r ♘a6?! , ANSWER: Yes, and absolutely typical of
( but the computer does not like this at all what happens in these structures, when they
f o r B l a c k , a n d p r e f e r s 17...♕b4 are handled well by such a great player as
, radically preventing the advance of the b- Bronstein. There are all manner of hidden
pawn. tactical chances and Black continually
QUESTION: That looks a bit dangerous manages to keep his d6-pawn alive by
after 18.♗d2 doesn't it? 19 Nd5 is a threat. exploiting them. Matthew Sadler once used
ANSWER: Indeed, but the silicon monster the analogy that playing White in such
assures us there is nothing to worry about positions is a bit like trying to squash a
after ♕a3 when White has nothing better jelly – you press down on one bit and it
than a repetition with 19.♗e3 ♕b4 , etc. pops up elsewhere. ]
QUESTION: So, the overall verdict is one 16...♘e5 Black keeps playing dynamically,
of equality, is it? pressuring c4 and threatening ...Nf3+.
ANSWER: Yes, I think that is fair. But [ By contrast, a passive defence of d6 with
the variations show how precise the two 16...♗f8? would be a clear mistake, and is
players must be, to maintain that balance almost always a bad sign in such
– even a small inaccuracy can be enough positions. ]
to leave a player suffering, especially 17.♘d4 The knight returns to d4, just two
Black. )] moves after leaving it – another confirmation
of what we have said earlier, that it is quite so his queen is free to move away after ...
hard for White to do without the knight on Nb4.
this square. Now d6 is no longer attacked, but, ♘b4 22.♕e2 ♗d7 QUESTION: How do
as Kasparov points out, Reshevsky was you assess the position? White seems to
probably hoping this would only be have been hassled a lot over the last few
temporary, and that he would be able to moves, but now things are stable, aren't they?
drive away the black knight with f4 and then ANSWER: Yes, that is true, but White still
return to the task of harassing d6. However, has to find something to do. Black intends ...
Black continues his guerrilla tactics. Re7 and ...Rae8. What is White to do?
a3! QUESTION: That is a bit of a surprise. QUESTION: Well, I would like to attack the
Doesn't this allow White to defend the c4- d6-pawn!
pawn securely with b3, without the a-file [ ANSWER: Naturally, but you will note
being opened or b3 itself becoming an object that, yet again, the knight on d4 cannot
of attack? m o v e a w a y : 22...♗d7 23.♘f3?? ( or
ANSWER: True, and it is noteworthy that 23.♘c2?? ♗xc3 24.♖xc3 ♘xe4 )
the computer does not particularly like this 23...♗xc3 24.♖xc3 ♘xe4 wins. ]
move. But Kasparov awards it an 23.e5?!
exclamation mark, and I agree. Effectively, [ White does not have anything particularly
Black is trading positional advantages – yes, c o n s t r u c t i v e t o d o , b u t t h e c o mp u t e r
he gives up his pressure against c4 and b3, suggests he should just wait, with a move
but in return, he establishes a blockade on such as 23.♗g2 ]
the dark squares c5 and b4, which will both [ or 23.♕f3 , when chances would be equal.
be occupied by the black knights. In addition, Instead, Reshevsky decides to get rid of his
the a3-pawn will be a trump card for Black e4 weakness, by trading it for Black's
in any endgame. weakness on d6. However, this backfires. ]
18.f4 ♘ed7 19.b3 ♘a6 QUESTION: 23...dxe5 24.fxe5 ♖ad8
Changing of the guard? [ T h e c o m p u t e r t h i n k s 24...♕c7
ANSWER: Indeed! This knight is heading to is even more accurate, but even after the
b4, and its colleague will take over the c5- text, the assessment of the position has
square. Note too, how the moves ...a3 and b3 changed.
have weakened the long diagonal. Now the QUESTION: Why?
white knight on d4 has to worry about its ANSWER: Because Black now simply
colleague on c3, which may be attacked if it stands better, as the pawn on e5 is weak,
moves away from d4. and White has no compensation for this. ]
QUESTION: So White still can't move his 25.g4 QUESTION: That looks rather
knight from d4, to attack d6? weakening!
ANSWER: No, or at least, it is not entirely ANSWER: It is, but White needs to get his
simple. bishop to h4 or g3.
20.♗f2 ♘dc5 21.♖e3 QUESTION: What is ♘e6 26.♗h4 ♘xd4 27.♖xd4 ♕c5!
this for? It looks a bit clumsy. As Bronstein notes, this is a key move.
ANS WER: I t d o es , but Wh ite needs t o Black must prevent the enemy knight
defend the c3-knight (see the previous note), reaching f6.
[ And, of course, not 27...♗xe5?? 28.♖xe5 ♖f8?! QUESTION: Why the punctuation?
♖xe5 29.♗xd8 , winning for White. ] ANSWER: This is an interesting moment.
28.♖de4 ♗h6! 29.♔h1 Both Bronstein and Kasparov award Black's
[ 29.♗xd8 ♖xd8 30.♕f2 ♗xe3 31.♖xe3 last move an exclamation mark, but Fritz
♕d4 32.♘e4 ♘xa2 is no better, ] claims it is a mistake, after which the
[ whilst 29.g5 ♗g7 is even worse, as the chances are equal.
black queen's bishop comes to f5 next [ The computer prefers 35...♖xe6 36.♖xe6
move. ] ♘b4 with an advantage. Note that
29...♗e6 30.g5 37.♖e8+? is then a mistake, since after
[ Kasparov describes this as a desperate ♖xe8 38.♕xe8+ ♕f8 White cannot defend
attempt to complicate, since the variation both f1 and a2. ]
30.♗xd8 ♖xd8 31.♖g3 (the point of 36.♖e7?
White's 29th move) ♖d2 32.♕e1 ♘c2 [ Fritz claims equal chances after 36.♗g2
33.♕b1 ♕f2 is hopeless. ] which Kasparov and Bronstein both ignore.
30...♗g7 31.♖f4 ♗f5 32.♘e4 Setting a very For example, ♘b4 37.♖e7 ♕d4 ( or
nasty trap. 37...♗d4? 38.♖f3 ♖xf3 39.♗xf3
♗xe4+! and the th reat o f Bg4-e6 is better for
[ 32...♕xe5?? loses to the vicious trick White ) 38.♖xb7 ♕xh4 39.♖xg7+ ♔xg7
33.♖xf5! gxf5 ( or 33...♕xf5 34.♘f6+ 40.♖e7+ ♖f7 41.♕e5+ ♔g8 42.♖xf7
♗xf6 35.♖xe8+ ♖xe8 36.♕xe8+ ♔g7 ♔xf7 43.♕f6+ ♔e8 44.♕e6+
37.gxf6+ ) 34.♘f6+ ♗xf6 35.gxf6!! , forcing perpetual check. ]
. The whole point. There is no defence 36...♗d4 37.♖3e6 Now we see the difference
against Qg2+. ] with the last note: White's pieces are hanging.
33.♖fxe4 ♘a6 QUESTION: Why is this [ 37.♖f3?? is impossible because of ♖xf3
strong? 38.♕xf3 ♕xe7 , whilst the text leads to
ANSWER: The knight heads to the problems down the f-file. ]
blockading square e6. Knights are well- [ 37.♖3e4 is similar to the game: ♕f5
known for being the ideal blockader. 38.♗g2 ♘c5 , and now 39.♖g4 ♕b1+
QUESTION: OK, but doesn't this allow 40.♔h2 ♕g1+ 41.♔g3 ♗f2+
White to offload his weak pawn? forces mate. ]
ANSWER: Yes, but, as will become apparent, 37...♕f5! 38.♖e8
this does not solve his problems. [ 38.♗g2 ♘c5 traps the rook on e6 in broad
34.e6 fxe6 35.♖xe6 QUESTION: So how daylight. ]
do you assess this position? 38...♘c5 39.♖xd8
ANSWER: Black has a clear advantage. [ 39.♖6e7 is elegantly refuted by ♖dxe8
Bronstein himself (who turned down a draw 40.♖xe8 ♘xb3! – another illustration of
offer in this position) identifies three the strength of the a3-pawn. ]
factors: the exposed white king, the 39...♘xe6 40.♖xf8+ ♔xf8 QUESTION:
superiority of the knight over the white Wouldn't the game have been adjourned
bishops, and the strength of the pawn on a3. around here?
The last factor was something we mentioned ANSWER: Yes, it was, with Reshevsky
in the note to Black's 17th move. sealing his next move. His position is almost
certainly lost, as the g5-pawn is weak, the 45.♔g2 ♕b2+ 46.♕e2+ ♔d6
white king exposed, and the a3-pawn still a [ The position after 46...♕xe2+? 47.♗xe2
permanent threat. However, Black's king is would be drawn, but instead, Black exploits
also exposed, and great care is required in the fact that White cannot himself exchange
converting the advantage. on b2, as the black king is too close. ]
41.♗g3 QUESTION: Giving up the g5-pawn 47.♔f3
at once? [ Kasparov quotes an old analysis by Euwe:
ANSWER: Yes, Reshevsky decides to 47.♕xb2? axb2 48.♗d3 ♔e5 49.♔f3 ♗c5
jettison the pawn immediately, in search of 50.♗b1 ♔f6 51.♔e4 h5 52.♗c2 g5
counterplay. There is already a trap. 53.♗b1 ♗d6 54.♗c2 g4 55.hxg4 hxg4
♕xg5! 56.♗b1 g3 57.♔f3 ♗f4 and, with the
[ 41...♘xg5? 42.♗d6+ ♔g7 43.♕e7+ white king tied down to covering the g-
♔h6 44.♕f8+ ♕xf8 45.♗xf8+ ♔h5 pawn, the black monarch just marches in on
46.♗xa3 is an immediate draw. ] the queenside. ]
42.♕xe6 ♕xg3 43.♕c8+ ♔e7 44.♕g4 47...♗c5 48.♔e4 ♕d4+ 49.♔f3 ♕f6+
QUESTION: Why not take b7 with check? 50.♔g2 ♔c7 51.♕f3 ♕b2+ 52.♕e2 ♕d4
[ ANSWER: Unfortunately, this is 53.♔f3 h5 QUESTION: It looks as though
i m p o s s i b l e , a s a f t e r 44.♕xb7+ ♔d8 Black has been groping for the right plan?
White runs out of checks within a few A N S W E R : W i t h a d j o u r n me n t a n a l y s i s
moves, and meanwhile, cannot stop the behind him, I rather doubt that. It was
mating threat: 45.♕a8+ ( or 45.♗g2 ♕e1+ always a feature of Soviet technique not to
46.♔h2 ♗e5# ) 45...♔c7! 46.♕a5+ ♗b6 rush such positions and to take one's time to
.] strengthen the position to the utmost, all the
44...♕c3 QUESTION: So Black is a pawn up, while the opponent has no counterplay. Now,
but we have opposite-coloured bishops. Is at any rate, Black commences the decisive
that not a big drawing factor? advance of the kingside pawns. This will
ANSWER: It is a drawing factor when there lead either to a mating attack or, after a
are no other pieces on the board, but with queen exchange, to a winning opposite-
queens on, as here, the tendency is if bishop ending, similar to that analysed in the
anything, the opposite – the possibility of a note to White's 47th.
attack, concentrated on the same colour 54.♔g2 g5 55.♔g3 ♕f4+ 56.♔g2 g4
squares as one's own bishop, is extremely 57.hxg4 hxg4 58.♔h1 ♔b6 59.♔g2 ♔c7
dangerous, since the opponent effectively 60.♔h1 ♗d6
has one fewer piece with which to defend [ In his analysis, Bronstein had found a neat
those squares. White's only drawing chances zugzwang finish, but the computer points
are either perpetual check, or else to out the much more prosaic and immediate
exchange queens at a favourable moment, 60...♕d4 61.♕g2 ♗d6 , and there is no
when the pure bishop ending will be drawn. defence to mate after ...Qh8. ]
However, the latter is not easy, as the 61.♔g1 ♔b6 62.♕g2 ♗c5+ 63.♔h1 ♕h6+
vulnerability of the a2-pawn to the black 64.♕h2 ♕e3 65.b4 ♗d4
king means that some of the bishop endings [ White is in zugzwang and lost on time. If
will be lost for White. 65...♗d4 66.♕g2 g3 67.c5+ ♔a7
[ QUESTION: OK, accepting what you say great David Bronstein, might have been
at face value for the moment, how does tempted to wimp out and play for a draw as
6.♘ge2 help White? Now he is not even White, but Malcolm has never been that type
winning a pawn after c5 . of player. He seizes the chance to take on his
ANSWER: No, indeed, but he intends to great opponent in a sharp and double-edged
meet this with 7.d5 , when he argues that fight. And quite right, too, regardless of the
he has a superior Benoni than would be the eventual result – after all, how often does
case with the knight on g1 and the bishop one get the chance to play against a
on e3. Once again, it is all fairly subtle challenger for the world championship?
modern theory, and is not really relevant ♕a5 10.♗h6?! Certainly consistent, but
to our present game, to which we now actually a novelty, according to the database.
return. ] [ In almost all the practical tests of this
6...e5 Bronstein was a much more classical position (and there are not many), White
KID player than many modern exponents of has preferred the prophylactic 10.♔b1
the opening, and he generally preferred ...e5 here, securing his a2-pawn and potentially
against the Sنmisch, in most variations. setting up Nd5 tactics against the
7.♗g5 This is another point of White's unguarded black queen.
choice at move 6. By delaying the QUESTION: But can Black continue as in
development of the bishop until Black has t h e g a m e w i t h b5 and be a tempo up? I
advanced his e-pawn, White is now able to thought in such opposite-castling positions,
bring the bishop to this more active square speed was of the essence?
than e3. The pin on the knight creates a few ANSWER: It is, but the rule of 'more haste,
problems for Black. less speed' also sometimes applies. The
c6 problem with 10...b5 in this position is that
[ QUESTION: Can't Black just kick the W h i t e h a s t h e s t r o n g r e p l y 11.♘d5!
bishop with 7...h6 ? , when after the almost forced line ♕xd2
ANSWER: He could, and this is probably 12.♘e7+ ♔h8 13.♖xd2 ♗b7 14.cxb5
the typical club player's reaction, but it cxb5 15.♘c3 he would have a nice
would be playing into White's hands by advantage. The b5-pawn is exposed, d5 is
weakening the black kingside. One must weak and White is well placed to continue
remember that White is still in a position with Be2 and Rhd1 (or Rc1), with the
to castle long and play for the attack on the advantage. ]
kingside, so such a weakening move as ... 10...b5?!
h6 could really come back to haunt Black. [ It looks even stronger to interpolate
After 8.♗e3 followed by Qd2 (with tempo) 10...♗xh6! 11.♕xh6 , dragging the white
and then 0-0-0 and h4 plus g4, White queen away, and only then to strike out
would have a very dangerous attack. ] with b5! . Now White has some problems,
8.♕d2 ♘bd7 9.0-0-0 White continues because 12...b4 is a threat, winning the a-
consistently, tucking his king away on the pawn, and this is not easy to meet:
queenside and preparing for a full-blooded 12.♔b1?? b4 wins the knight, a line which
struggle. One must admire Malcolm Pein for shows one of the problems of having the
his approach here – many players, facing the king's knight on e2. ]